ViewChange
Harnessing the power of storytelling to change the world™

Entity Extraction & Content API Evaluation

Posted on May 18th, 2010 by Rob DiCiuccio

The ViewChange.org platform utilizes several third-party APIs to perform semantic content analysis and related content aggregation. The technologies developed by these partners have allowed us to leverage powerful tools for entity extraction and content discovery, and facilitate integration with the emerging Semantic Web. These tools allow us to focus more of our efforts on creating an engaging user experience, as well as building new open source tools for content curation, without duplicating the efforts of others.

There are several open APIs that provide semantic analysis and content discovery services. The following evaluation was conducted in February 2010, in order to assess the available APIs and their compatibility with the needs of the ViewChange.org platform. The API tests were performed using a custom PHP application for querying and displaying API results. If you would like the source for the test application, send me a message.


  1. Overview
  2. Test Data
    1. Segment Transcript #1
    2. Segment Transcript #2
    3. Segment Transcript #3
  3. Natural Language Processing (NLP) & Entity Extraction (EE) APIs
    1. OpenCalais
    2. Zemanta
    3. AlchemyAPI
    4. Evri
    5. OpenAmplify
    6. Yahoo! Term Extraction
  4. Content APIs – Articles
    1. Zemanta
    2. Daylife
    3. Bing API 2.0 (News Search)
    4. Yahoo! BOSS (News Search)
  5. Content APIs – Videos
    1. YouTube
    2. Truveo
    3. Bing API 2.0 (Video Search)
    4. Vimeo
  6. Content APIs – Actions
    1. SocialActions
  7. Other APIs & Partners
    1. Freebase
    2. DBpedia
  8. Conclusions & Recommendations
    1. Entity Extraction APIs
    2. Content APIs – Articles
    3. Content APIs – Videos
    4. Content APIs – Actions

Overview

The purpose of these tests is to evaluate available APIs for Entity Extraction and related content, as it pertains to the ViewChange.org initiative of Link Media. The tests will focus on the quality of the results, in terms of relevance and completeness of metadata returned, as well as the features and flexibility of each API. Sample text from ViewChange video segments, as well as Link TV content, will be used as the test data. The test data text segments (below) will serve as the basis for extracting Named Entities, which, in turn, will be used to query related content APIs.

Test Data

Segment Transcript #1

The area here is called Joubert Park, it is very close to Hillbrow, which is the largest residential place in the city of Johannesburg. It is a highly transitory place. It is a place where people arrive into South Africa. Many people from outside the country, from African countries, neighboring countries, as well as from overseas, other continents as well as from other provinces, parts of Africa that are mostly remote and rural, they arrive in Johannesburg to here.

Segment Transcript #2

What we are doing here at the GreenHouse is empowering the people so they can realize that they’ve got all the knowledge. Many of them here come from rural communities. They have once lived like this. They have once produced their own food. They have once built their own houses. They have once fetched their own water. They have once dealt with their own waste. The GreenHouse project has 5 programs and the first is green building and design where we focus on buildings and how we design our own buildings with positive solar design to minimize the opportunity to warm them up and cool them off with different kinds of materials. [Dorah Lebelo: Director – The GreenHouse Project] The second program is efficient renewable energy. What are the newest options that are available to replace the coal based electricity? The recycling project was started about 2.5 years ago, mainly just to showcase that waste is a decent resource that can be utilized. The fifth one is organic food production and nutrition, so we are looking at how people can start growing their own food in the city. We are looking at the principle of doing more with less, from a place of abundance, knowing that we’ve got what we need. We’ve got everything that we need. We are not going to look at other people to give us what we need. We’ve got what we need. We want the people to maximize our potential. If we want to create sustainable communities, we are going to have to look at things in a holistic way. We just can’t come and say “My responsibility is health and I’m just going to…I’m not going to come here and only look at health, and I’m just going to give these people drugs and help them survive AIDS.” You need to look at what it is they are eating, where they are living, what kind of houses are they living in. What kind of energy are they using, because if they are using coal and they are inhaling the smoke at night, it’s not going to be helpful. Its not only about one thing, it is about a number of things and most of them have a local effect, one thing leading to another. How do we design interventions, programs that are looking at the lives of the people in a holistic way rather than just one thing?

Segment Transcript #3

The Goldstone Report accuses Israel and Hamas of war crimes. Israeli officials are furious and reject the findings. Why is Arab media covering up for Hamas? Is Israel above the law?

Natural Language Processing (NLP) & Entity Extraction (EE) APIs

OpenCalais

Overview
A product of Thomson Reuters, OpenCalais provides a robust Natural Language Processing engine to extract semantic entities from text. The open API has been widely adopted by the Open Source community, powering the OpenPublish platform and integrating with Drupal and WordPress. OpenCalais is one of the big players in the NLP space, providing semantic metadata for several high–volume sites, such as The Huffington Post and CNET.

Service Description

Notes

  1. Currently provides entity disambiguation for companies, geographies and electronic products only
  2. Entity URIs contain “/er/” to signify disambiguation, e.g. “http://d.opencalais.com/er/geo/city/ralg–geo1/3d7ed876–a27–df3–9b8–7d1d73a90e5″
  3. Supports entry of text documents up to 100K characters in length
  4. Development seems to be primarily business/commerce oriented
  5. Provides additional abstract “Social Tags” and “Facts and Events” results, though they are not linked to other databases such as FreeBase

Analysis
The entities provided by the Calais Web Service are relevant, and generally of good quality. What is lacking, however, is entity disambiguation and connections to Linking Open Data (LOD) datasets. OpenCalais currently provides entity disambiguation and Linked Data URIs for only a small subset of entity types. Other NLP APIs offer more comprehensive entity disambiguation and Linked Data features. While OpenCalais does provide additional relevant “Social Tags” and “Facts and Events” metadata, these results are somewhat proprietary, and are not dereferenceable or linked to other entity databases.

Results: Segment Transcript #1
Name: Johannesburg [score: 0.714]
Linked Data:

  • http://d.opencalais.com/er/geo/city/ralg–geo1/3d7ed876–a27–df3–9b8–7d1d73a90e5

Name: Africa [score: 0.238]
Linked Data:
Name: South Africa [score: 0.365]
Linked Data:

  • http://d.opencalais.com/er/geo/country/ralg–geo1/c814068b–bc92–07ac–d130–de362b2d6ddc

Query took 0.633252859116 seconds

Note: also returned ambiguous “Topic” results: “Environment,” “Hospitality_Recreation”

Results: Segment Transcript #2
Name: organic food production [score: 0.182]
Linked Data:

Name: electricity [score: 0.26]
Linked Data:

Name: Dorah Lebelo [score: 0.333]
Linked Data:

Name: GreenHouse Project [score: 0.333]
Linked Data:

Name: Director [score: 0.333]
Linked Data:

Name: renewable energy [score: 0.361]
Linked Data:

Name: food [score: 0.599]
Linked Data:

Name: energy [score: 0.125]
Linked Data:

Query took 3.59928512573 seconds

Note: also returned ambiguous “Topic” results: “Environment”

Results: Segment Transcript #3
Name: Hamas [score: 0.639]
Linked Data:

Name: Israel [score: 0.714]
Linked Data:

  • http://d.opencalais.com/er/geo/country/ralg–geo1/3f8454b7–f2c1–2ceb–b162–2b6b0dfd1021

Query took 0.695325136185 seconds

Note: also returned ambiguous “Topic” results: “Politics,” “War_Conflict,” “Law_Crime”

Zemanta

Overview
Primarily marketed as a blog enhancement product, Zemanta provides an open API for its semantic entity extraction and related content services. Zemanta is a Slovenia–based start–up, launched in 2007, and is backed by Accelerator Group, Britain’s Eden Ventures, and Union Square Ventures.

Service Description

Notes

  • Per docs: “Only first 8kb of text is going to be processed
  • Provides additional ambiguous “keywords” results when using the zemanta.suggest method
  • May be possible to enhance results using the “return_categories” with partner ID feature (need to inquire further)

Analysis
Zemanta is a unique product, in that it offers both semantic entity extraction and related content in a single API call. The entities returned via the zemanta.suggest service are disambiguated and provide several links to open databases such as FreeBase and DBpedia. The entity results range from highly specific topics to more general concepts and categories, presented on a sliding relevance scale. Overall, the entity results are highly relevant, and include more than enough Linked Data information to make informed editorial decisions about content/entity relationships.

Results: Segment Transcript #1
Name: Africa [score: 0.565062]
Linked Data:

  • http://en.wikipedia.org/wiki/Africa
  • http://rdf.freebase.com/ns/en/africa
  • http://dbpedia.org/resource/Africa

Name: Hillbrow [score: 0.514032]
Linked Data:

  • http://maps.google.com/maps?ll=–26.1888888889,28.0491666667&spn=0.01,0.01&q=–26.1888888889,28.0491666667 (Hillbrow)&t=h
  • http://en.wikipedia.org/wiki/Hillbrow
  • http://rdf.freebase.com/ns/en/hillbrow_gauteng
  • http://dbpedia.org/resource/Hillbrow

Name: Johannesburg [score: 0.504859]
Linked Data:

  • http://maps.google.com/maps?ll=–26.2044444444,28.0455555556&spn=0.1,0.1&q=–26.2044444444,28.0455555556 (Johannesburg)&t=h
  • http://www.joburg.org.za/
  • http://en.wikipedia.org/wiki/Johannesburg
  • http://rdf.freebase.com/ns/en/johannesburg
  • http://dbpedia.org/resource/Johannesburg

Name: rural [score: 0.428101]
Linked Data:

  • http://en.wikipedia.org/wiki/Rural_area
  • http://rdf.freebase.com/ns/en/rural
  • http://dbpedia.org/resource/Rural_area

Name: countries [score: 0.408625]
Linked Data:

  • http://en.wikipedia.org/wiki/Country
  • http://rdf.freebase.com/ns/en/country
  • http://dbpedia.org/resource/Country

Name: continents [score: 0.375792]
Linked Data:

  • http://en.wikipedia.org/wiki/Continent
  • http://rdf.freebase.com/ns/en/continent
  • http://dbpedia.org/resource/Continent

Query took 1.6244790554 seconds

Note: zemanta.suggest also returned ambiguous “keywords” results: South Africa, Africa, Johannesburg, Hillbrow, Rural area, Government, Travel and Tourism, Travel Guides

Results: Segment Transcript #2
Name: green building [score: 0.574009]
Linked Data:

  • http://en.wikipedia.org/wiki/Green_building
  • http://rdf.freebase.com/ns/en/green_building
  • http://dbpedia.org/resource/Green_building

Name: organic food [score: 0.549828]
Linked Data:

  • http://en.wikipedia.org/wiki/Organic_food
  • http://rdf.freebase.com/ns/en/organic_food
  • http://dbpedia.org/resource/Organic_food

Name: renewable energy [score: 0.543073]
Linked Data:

  • http://www.wikinvest.com/industry/Renewable_Energy
  • http://en.wikipedia.org/wiki/Renewable_energy
  • http://rdf.freebase.com/ns/en/renewable_energy
  • http://dbpedia.org/resource/Renewable_energy

Name: AIDS [score: 0.483078]
Linked Data:

  • http://en.wikipedia.org/wiki/AIDS
  • http://rdf.freebase.com/ns/en/aids
  • http://dbpedia.org/resource/AIDS

Name: solar design [score: 0.46359]
Linked Data:

  • http://en.wikipedia.org/wiki/Passive_solar_building_design
  • http://rdf.freebase.com/ns/en/passive_solar_building_design
  • http://dbpedia.org/resource/Passive_solar_building_design

Name: recycling [score: 0.45182]
Linked Data:

  • http://en.wikipedia.org/wiki/Recycling
  • http://rdf.freebase.com/ns/en/recycling
  • http://dbpedia.org/resource/Recycling

Name: sustainable communities [score: 0.446424]
Linked Data:

  • http://en.wikipedia.org/wiki/Sustainable_community
  • http://rdf.freebase.com/ns/en/sustainable_community
  • http://dbpedia.org/resource/Sustainable_community

Name: water [score: 0.430005]
Linked Data:

  • http://en.wikipedia.org/wiki/Water
  • http://rdf.freebase.com/ns/en/water
  • http://dbpedia.org/resource/Water

Name: electricity [score: 0.427965]
Linked Data:

  • http://en.wikipedia.org/wiki/Electricity
  • http://rdf.freebase.com/ns/en/electricity
  • http://dbpedia.org/resource/Electricity

Name: energy [score: 0.422151]
Linked Data:

  • http://www.wikinvest.com/industry/Energy
  • http://en.wikipedia.org/wiki/Energy
  • http://rdf.freebase.com/ns/en/energy
  • http://dbpedia.org/resource/Energy

Query took 3.76841306686 seconds
Note: zemanta.suggest also returned ambiguous “keywords” results: Energy, Renewable energy, Green building, Business, Organic food, Recycling, Technology

Results: Segment Transcript #3
Name: Goldstone Report [score: 0.580401]
Linked Data:

  • http://en.wikipedia.org/wiki/United_Nations_Fact_Finding_Mission_on_the_Gaza_Conflict
  • http://www.youtube.com/watch?v=8vZeBbLeo–M
  • http://www.youtube.com/watch?v=XtbHifKM6sM
  • http://www.youtube.com/AlJazeeraEnglish
  • http://www.youtube.com/watch?v=pp9B7p9AFwY
  • http://dbpedia.org/resource/United_Nations_Fact_Finding_Mission_on_the_Gaza_Conflict

Name: Israel [score: 0.501919]
Linked Data:

  • http://maps.google.com/maps?ll=31.7833333333,35.2166666667&spn=1.0,1.0&q=31.7833333333,35.2166666667 (Israel)&t=h
  • http://en.wikipedia.org/wiki/Israel
  • http://rdf.freebase.com/ns/en/israel
  • http://dbpedia.org/resource/Israel

Name: Hamas [score: 0.489207]
Linked Data:

  • http://en.wikipedia.org/wiki/Hamas
  • http://www.youtube.com/watch?v=O8TTjb54GzM
  • http://www.youtube.com/watch?v=LCVr7MBhgj0
  • http://rdf.freebase.com/ns/en/hamas
  • http://dbpedia.org/resource/Hamas

Query took 1.0125310421 seconds

Note: zemanta.suggest also returned ambiguous “keywords” results: Hamas, Israel, United Nations Fact Finding Mission on the Gaza Conflict, Middle East, War crime, United Nations, Israel Defense Forces, Warfare and Conflict

AlchemyAPI

Overview
AlchemyAPI provides Named Entity Extraction and disambiguation capabilities for analyzing text, HTML or scanned images. Additional features include language identification, quotation extraction and content scraping & structured data extraction. Created by Orchestr8, in operation since 2005.

Service Description

Notes

  • Seems to be missing disambiguated URIs for “person” entities.
  • Very fast response time (under 1s).

Analysis
Overall, the entity quality is good, but the number of entities returned is low. Returned entities are, with a few exceptions, disambiguated and provide dereferenceable links to other structured databases such as Freebase and DBpedia. Calls to the API are very fast, and the usage limits are generous.

Results: Segment Transcript #1
Name: Johannesburg [score: 0.883671]
Linked Data:

  • –26.2044444444 28.0455555556 [geo]
  • http://www.joburg.org.za/ [website]
  • http://dbpedia.org/resource/Johannesburg [dbpedia]
  • http://rdf.freebase.com/ns/guid.9202a8c04000641f8000000000070904 [freebase]
  • http://sws.geonames.org/993800/ [geonames]
  • http://sw.opencyc.org/concept/Mx4rvViMQpwpEbGdrcN5Y29ycA [opencyc]

Name: Joubert Park [score: 0.734335]
Linked Data:
Name: South Africa [score: 0.730116]
Linked Data:

  • http://dbpedia.org/resource/South_Africa [dbpedia]
  • http://rdf.freebase.com/ns/guid.9202a8c04000641f800000000007fa5e [freebase]
  • http://www4.wiwiss.fu–berlin.de/factbook/resource/South_Africa [ciaFactbook]
  • http://umbel.org/umbel/ne/wikipedia/South_Africa [umbel]
  • http://sw.opencyc.org/concept/Mx4rvVkJaJwpEbGdrcN5Y29ycA [opencyc]
  • http://mpii.de/yago/resource/South_Africa [yago]

Name: Africa [score: 0.695728]
Linked Data:

  • http://dbpedia.org/resource/Africa [dbpedia]
  • http://rdf.freebase.com/ns/guid.9202a8c04000641f8000000000c70e81 [freebase]
  • http://umbel.org/umbel/ne/wikipedia/Africa [umbel]
  • http://sw.opencyc.org/concept/Mx4rvVjtJ5wpEbGdrcN5Y29ycA [opencyc]
  • http://mpii.de/yago/resource/Africa [yago]

Query took 0.232305049896 seconds

Results: Segment Transcript #2
Name: AIDS [score: 0.946591]
Linked Data:

  • Disease [subType]
  • http://dbpedia.org/resource/AIDS [dbpedia]
  • http://rdf.freebase.com/ns/guid.9202a8c04000641f8000000000c0a7a2 [freebase]
  • http://umbel.org/umbel/ne/wikipedia/AIDS [umbel]
  • http://sw.opencyc.org/concept/Mx4rwP1LGpwpEbGdrcN5Y29ycA [opencyc]
  • http://mpii.de/yago/resource/AIDS [yago]

Name: Dorah Lebelo [score: 0.942957]
Linked Data:

Query took 0.646384000778 seconds

Results: Segment Transcript #3
Name: Israel [score: 0.915]
Linked Data:

  • http://dbpedia.org/resource/Israel [dbpedia]
  • http://rdf.freebase.com/ns/guid.9202a8c04000641f800000000001e2be [freebase]
  • http://www4.wiwiss.fu–berlin.de/factbook/resource/Israel [ciaFactbook]
  • http://umbel.org/umbel/ne/wikipedia/Israel [umbel]
  • http://sw.opencyc.org/concept/Mx4rvVjqYZwpEbGdrcN5Y29ycA [opencyc]
  • http://mpii.de/yago/resource/Israel [yago]

Name: Hamas [score: 0.910833]
Linked Data:

  • http://www.palestine–info.com/ [website]
  • http://dbpedia.org/resource/Hamas [dbpedia]
  • http://rdf.freebase.com/ns/guid.9202a8c04000641f800000000001ccd0 [freebase]
  • http://umbel.org/umbel/ne/wikipedia/Hamas [umbel]
  • http://mpii.de/yago/resource/Hamas [yago]

Query took 0.219098091125 seconds

Evri

Overview
Evri provides several APIs for NLP text analysis, content recommendations and relationships between semantic entities. The focus here seems to be on relationships between entities and related content, as well as topic popularity.

Service Description

Notes

  • Some problems with availability and returning empty results
  • Caches query results based on submitted ‘uri’ parameter value

Analysis
Entity extraction from submitted text returns (mostly) disambiguated Evri identifiers, but no Linked Data URIs to other databases. The quality of the entities is somewhat hit–and–miss, and seems to be lacking in general concept results, when compared with other NLP APIs. Requires multiple queries to various Evri API endpoints to retrieve factual information & links related to the entity. Terms of Service seems fairly restrictive regarding reordering results, etc.

Results: Segment Transcript #1
Name: Johannesburg [score: 85.0]
Linked Data:

  • http://www.evri.com/location/johannesburg–0×1172e6

Name: Joubert Park [score: 43.0]
Linked Data:

  • http://www.evri.com/location/joubert–park

Name: South Africa [score: 35.0]
Linked Data:

  • http://www.evri.com/location/south–africa–0×32317

Name: Africa [score: 27.0]
Linked Data:

  • http://www.evri.com/location/africa–0×4c5b0

Query took 0.156415939331 seconds

Results: Segment Transcript #2
Name: Coal [score: 102.0]
Linked Data:

  • http://www.evri.com/substance/coal–0×397297

Name: Waste [score: 102.0]
Linked Data:

  • http://www.evri.com/concept/waste–0×395eea

Name: AIDS [score: 51.0]
Linked Data:

  • http://www.evri.com/condition/aids–0xef0c7

Query took 0.563608169556 seconds

Results: Segment Transcript #3
Name: Hamas [score: 85.0]
Linked Data:

  • http://www.evri.com/organization/hamas–0×346b1

Name: Israel [score: 38.0]
Linked Data:

  • http://www.evri.com/location/israel–0×15f8d

Query took 0.178416967392 seconds

OpenAmplify

Overview
OpenAmplify provides Natural Language Processing APIs, for use in commercial applications. According to the OpenAmplify website, they have “concentrated on commercial systems that meet the demands of large–scale business use.”

Service Description

Notes

  • Initial attempts to use the API resulted in connection timeouts, subsequent requests were very slow (15+ seconds)
  • API URL was changed during testing to mitigate connection problems

Analysis
The entities returned from the OpenAmplify NLP API are generally of poor quality, based on these tests. Abstract terms and seemingly irrelevant words are given high scores, and phrases such as “Goldstone Report” are broken into single word results that have no meaning when separated. Topics are returned along with ambiguous category information (e.g. “Healthcare, Disease, Nutrition”) as well as topic “polarity” (negative/positive) and intentions. Returned values are not disambiguated and have no Linked Data URIs.

Results: Segment Transcript #1
Name: people [score: 10]
Linked Data:

Name: country [score: 9]
Linked Data:

Name: Johannesburg [score: 8]
Linked Data:

Name: Joubert [score: 7]
Linked Data:

Name: park [score: 7]
Linked Data:

Query took 0.560339927673 seconds

Results: Segment Transcript #2
Name: people [score: 24]
Linked Data:

Name: GreenHouse [score: 18]
Linked Data:

Name: program [score: 15]
Linked Data:

Name: project [score: 14]
Linked Data:

Name: coal [score: 10]
Linked Data:

Query took 6.6040558815 seconds

Results: Segment Transcript #3
Name: Israel [score: 14]
Linked Data:

Name: Hamas [score: 11]
Linked Data:

Name: Goldstone [score: 7]
Linked Data:

Name: report [score: 7]
Linked Data:

Name: finding [score: 5]
Linked Data:

Query took 1.8968129158 seconds

Yahoo! Term Extraction

Overview
Part of the Yahoo! Developer Network, the Term Extraction Web Service is a bare–bones term extraction API.

Service Description

Analysis
The Yahoo! Term Extraction API is not a semantic entity extraction tool. It returns words or phrases with no disambiguation or Linked Data information. It appears that the terms originate from Yahoo’s search index, though there is little documentation to this effect.

Results: Segment Transcript #1
Name: city of johannesburg [score: ]
Linked Data:

Name: hillbrow [score: ]
Linked Data:

Name: place where people [score: ]
Linked Data:

Name: african countries [score: ]
Linked Data:

Name: continents [score: ]
Linked Data:

Name: south africa [score: ]
Linked Data:

Query took 0.135603189468 seconds

Results: Segment Transcript #2
Name: organic food production [score: ]
Linked Data:

Name: greenhouse project [score: ]
Linked Data:

Name: recycling project [score: ]
Linked Data:

Name: solar design [score: ]
Linked Data:

Name: sustainable communities [score: ]
Linked Data:

Name: rural communities [score: ]
Linked Data:

Name: different kinds [score: ]
Linked Data:

Name: renewable energy [score: ]
Linked Data:

Name: coal [score: ]
Linked Data:

Name: abundance [score: ]
Linked Data:

Name: electricity [score: ]
Linked Data:

Name: principle [score: ]
Linked Data:

Name: showcase [score: ]
Linked Data:

Name: aids [score: ]
Linked Data:

Name: drugs [score: ]
Linked Data:

Name: nutrition [score: ]
Linked Data:

Name: health [score: ]
Linked Data:

Query took 0.584623098373 seconds

Results: Segment Transcript #3

Name: war crimes [score: ]
Linked Data:

Name: arab media [score: ]
Linked Data:

Name: israeli officials [score: ]
Linked Data:

Name: hamas [score: ]
Linked Data:

Name: israel [score: ]
Linked Data:

Query took 0.267089128494 seconds

Content APIs – Articles

Test results, in most cases, are using entities returned from OpenCalais or Zemanta’s entity extraction engine, and have not been curated. In the application environment, these entities will be curated by content managers, so the quality of the results below are not entirely accurate; however, they do demonstrate the types of metadata returned from each API. The first five article results are shown for each query.

Zemanta

Overview
Zemanta’s content API (zemanta.suggest) provides related articles and images. See company description above.

Service Description
See company description above

Notes

  • Using transcript text (first 8kb), not extracted entities, for related content analysis.
  • Extracted entities and related content can be returned in a single query.
  • No description or article abstract text returned [UPDATE: article excerpt text is returned when setting the undocumented articles_highlight parameter to 1]
  • No data returned for the article’s source, other than what can be extrapolated from the article’s URL
  • Some article URLs point to a redirection script on Zemanta’s site, though the original article URL can be parsed out
  • Possible to influence results using the “emphasis” parameter, though more information is needed here
  • Possible to limit the content results to specific sources by using the ’sourcefeed_ids’ parameter, though the IDs are not publicly available

Analysis
Zemanta provides a unique content API, due to the fact that it provides related content based on submitted text blocks, instead of individual keywords or search terms. As with Zemanta’s suggest_markup method, NLP entity extraction is performed on the text to identify contextually relevant topics, then those topics are used to return related articles & blog posts. The article results in these tests are somewhat across the board, though the results are both relevant and timely when specific events appear within the text. The downside to the Zemanta content API is the limited metadata returned for each item, including description text and source information [UPDATE: see above].

Results: Segment Transcript #1

Name: Jacob Zuma acknowledges fathering love child [score: 0.003537]
Publish Date: 2010–02–03T19:48:51Z
Description:
Source:

Name: How does a gender lens contribute to a healthy environment and healthy people? Intersections between gender, HIV and migration: Lessons from South Africa [score: 0.00276]
Publish Date: 2010–01–06T13:06:37Z
Description:
Source:

Name: Steven Pienaar dreams of walking out for South Africa in the World Cup’s first game [score: 0.002305]
Publish Date: 2009–09–12T21:34:59Z
Description:
Source:

Name: Gee–raff – still chasing the Bang Bang [score: 0.002047]
Publish Date: 2009–11–04T12:57:27Z
Description:
Source:

Name: S.Africa: No reply on terror query, frees Tunisian [score: 0.001896]
Publish Date: 2010–01–27T17:35:14Z
Description:
Source:

Query took 1.64199209213 seconds

Results: Segment Transcript #2

Name: IES Appoints New Canadian Energy Modelling Consultant and Invests Further in Software Development [score: 0.00365]
Publish Date: 2010–02–03T15:04:00Z
Description:
Source:

Name: Horizon at Playa Vista Becomes First Ground–Up Office Building Construction in Los Angeles to Receive Prestigious Gold LEED Certification in ‘Core & Shell’ Category [score: 0.002714]
Publish Date: 2010–02–01T17:03:00Z
Description:
Source:

Name: Intel plans 8 solar arrays to help power facilities [score: 0.002473]
Publish Date: 2010–02–01T14:08:09Z
Description:
Source:

Name: Green Building Final Presentation [score: 0.00241]
Publish Date: 2010–01–29T21:49:52Z
Description:
Source:

Name: LEED Gold Certification for Cape Cod Green Home Achieved by Cape Associates and ZeroEnergy Design [score: 0.002207]
Publish Date: 2010–02–02T15:00:00Z
Description:
Source:

Query took 3.74818515778 seconds

Results: Segment Transcript #3
Name: Israel Defends Its Inquiry Into Gaza War [score: 0.020868]
Publish Date: 2010–01–30T01:25:21Z
Description:
Source:

Name: Israeli officer says rules of engagement ignored [score: 0.020634]
Publish Date: 2010–02–03T16:58:33Z
Description:
Source:

Name: Alan Dershowitz: Israel’s Military Investigation: Is it Enough? [score: 0.020263]
Publish Date: 2010–02–02T23:24:57Z
Description:
Source:

Name: Israel submits Goldstone response [score: 0.019466]
Publish Date: 2010–01–29T14:50:54Z
Description:
Source:

Name: Israel to release report into Gaza offensive [score: 0.019343]
Publish Date: 2010–01–29T14:30:10Z
Description:
Source:

Query took 1.03602385521 seconds

Daylife

Overview
The Daylife API provides timely news & blog content, as well as quotes and images, based on submitted topics. Daylife provides content services to several large news organizations, including USA Today and the Washington Post.

Service Description

Notes

  • Per docs : Search News Object can accept queries that are composed of up to 32 terms. Each term is considered to be a word where space is a word boundary. The logical operators are not counted in these 32 terms.
  • Supports weighting or “boosting” of individual term relevance
  • Supports logical operators OR, AND, NOT, as well as grouping via parentheses
  • Supports source filtering for content source whitelists and blacklists, as well as predefined filters
  • API called can be grouped into “batch API” calls, reducing the number of queries required for multiple content types

Analysis
What sets the Daylife API apart from the other content APIs compared here is the ability to “boost ” search terms using a numeric weight value. This allows multiple search terms to be prioritized in order of relevance. Daylife also provides robust metadata for each content item returned, including full source data and excerpt text. Daylife claims to have over 7000 sources feeding its content database, which is demonstrated by the occasionally obscure sources in the results. Overall, the Daylife API provides a high level of flexibility, and high quality results. The only potential downside seems to be the query limit, which is a bit low.

Results: Segment Transcript #1 (OpenCalais entities)

Name: D.C. United Academy Returns From South Africa [score: 10]
Publish Date: 2010-02-03 22:17:45
Description: …tournament, the Black-and-Red posted a 3-2 overall record, defeating clubs from France, India, and South Africa. In the team’s final match of the competition United defeated SAFA Johannesburg 4-2 on penalties, capping the experience of a lifetime for both…
Source: Our Sports Central

Name: Lopez lands in Johannesburg QFs [score: 9.68048132242]
Publish Date: 2010-02-03 14:18:36
Description: …South Africa (Sports Network) – Third-seeded Feliciano Lopez of Spain highlighted Wednesday’s second-round winners at the $500,000 SA Tennis Open. The lefthanded Lopez got past Slovenian Blaz Kavcic 6-2, 7-6 (7-4) on the hardcourts at Montecasino. Lopez’ quarterfinal…
Source: TSN.ca

Name: Ram, Robert advance in Johannesburg [score: 9.47635429611]
Publish Date: 2010-02-02 04:14:58
Description: …South Africa — Fifth-seeded Rajeev Ram of the United States and No. 8 Stephane Robert of France were among first-round winners Monday in the SA Tennis Open in Johannesburg. Ram came back after dropping the opening set and defeated Rik De Voest, a wild-card…
Source: The Money Times

Name: South Africa: Ecobank Opens Representative in Country [score: 9.47390085192]
Publish Date: 2010-02-01 23:35:03
Description: …has opened a representative office in Johannesburg, making South Africa the 30th African country in which the pan-African bank has a presence. The South African office will enable Ecobank to serve a growing number of its customers currently doing business…
Source: AllAfrica.com

Name: South Africa: Mozambican Airline in SA Airlink Deal [score: 9.27996692035]
Publish Date: 2010-02-03 11:08:21
Description: …several European airlines for a possible partnership but their management believed that partnering with a South African airline made more sense. Foster believes the fast-developing nation provides the new airline with huge scope to grow. Copyright © 2010 Business…
Source: AllAfrica.com

Query took 0.196148872375 seconds

Results: Segment Transcript #1 (Zemanta entities)

Name: South Africa: Mozambican Airline in SA Airlink Deal [score: 10]
Publish Date: 2010-02-03 11:08:21
Description: …between the two countries was expanded to include a second carrier from both countries. Previously, the Johannesburg-Maputo route was served only by LAM, Mozambique’s flag carrier, but with the changes to the bilateral agreement in July last year, 1time and…
Source: AllAfrica.com

Name: South Africa: Ecobank Opens Representative in Country [score: 9.14387846505]
Publish Date: 2010-02-01 23:35:03
Description: …skills and competences to bear on banking businesses in the eastern and southern African sub-regions. Our South Africa Representative Office will be an anchor point in this drive.” In December 2008, Ecobank entered into a strategic business cooperation alliance…
Source: AllAfrica.com

Name: Africa: AU Displays a Rare Glimpse of Courage [score: 8.07880556962]
Publish Date: 2010-02-02 11:34:10
Description: …— AFRICAN leaders displayed rare courage at the African Union (AU) meeting this weekend when they thwarted an attempt by Libyan leader Muammar Gaddafi to extend his chairmanship of the organisation for a second term. The row over the possible extension of…
Source: AllAfrica.com

Name: Impoverished African countries pledge aid to Haiti, but can they afford it? [score: 8.06276539473]
Publish Date: 2010-02-03 20:24:02
Description: …should not prevent us from helping a brother country.” The notion of a common African brotherhood that extends to the Caribbean and Americas remains a powerful symbol here on the African continent. Called Pan-Africanism by some and “negritude” by liberation…
Source: McClatchy

Name: Tunisia: Country Tops Electrification Rate in Africa [score: 7.98847356002]
Publish Date: 2010-02-02 11:39:41
Description: …power sufficiency and economic development is unmistakable.” Noting that some African countries are approaching universal electrification, the report adds that in Tunisia, Libya, Algeria and Egypt there has been a “sustained government support for rural electrification…
Source: AllAfrica.com

Query took 0.441734790802 seconds

Results: Segment Transcript #2 (OpenCalais entities)

Name: European Firms Snapping Up US Wind Resources [score: 10]
Publish Date: 2010-02-01 17:44:02
Description: …Administration predicted how the world will look in the year 2035 if we simply continue with the energy policies we have in place today: The amount of electricity generated from coal will remain above 40 percent, the amount generated from renewables will be…
Source: CNBC

Name: Landowners dismayed over renewable energy drive [score: 8.99741638643]
Publish Date: 2010-02-02 11:55:34
Description: …scheme was still only aiming to deliver just two per cent of the UK’s energy from onsite renewables by 2020. “It is good news for the domestic sector and is much better than what we had before, but it is much less encouraging for businesses,” she said. Friends…
Source: SouthWestBusiness.co.uk

Name: Firm plans Milwaukee renewable energy plant [score: 8.70462676457]
Publish Date: 2010-02-02 21:29:35
Description: …and industrial wastes into electricity. The firm said its “Project Apollo” will use Westinghouse Plasma Corp.’s plasma gasification technology for its 25-megawatt renewable energy facility. The project is expected to create 250 construction jobs and 45 full-time…
Source: Business Journal of Milwaukee

Name: Underscoring its Commitment to Cleaner Energy Solutions, GE Dedicates $45 Million, Eco-Friendly Renewable Energy Global Headquarters [score: 8.06378439597]
Publish Date: 2010-02-03 04:36:06
Description: …$6 billion.” The event also marked the installation of GE’s 13,500th wind turbine globally, further demonstrating the continued growth of GE’s renewable energy business. GE is the largest supplier of wind turbines in North America, and the company’s 1.5-megawatt…
Source: TheAutoChannel.com

Name: Recommendations to ensure future of UK renewable energy [score: 7.69133875549]
Publish Date: 2010-02-03 14:08:45
Description: …and electricity watchdog Ofgem has proposed an end to the policy of Renewables Obligation in order to encourage future investment in renewable energy. The proposal comes as part of a raft of recommendations announced by the regulator to ensure Britain’s energy…
Source: Materials Recycling Week

Query took 0.751944065094 seconds

Name: Underscoring its Commitment to Cleaner Energy Solutions, GE Dedicates $45 Million, Eco-Friendly Renewable Energy Global Headquarters [score: 10]
Publish Date: 2010-02-03 04:36:06
Description: …Energy and Environmental Design (LEED) green building standards and will be 20% more energy efficient than required by New York State building standards. Features include low-water faucets and improved insulation, energy efficient hot water boilers and air…
Source: TheAutoChannel.com

Name: Kroon Hall receives the U.S. Green Building Council’s highest rating [score: 8.53270566029]
Publish Date: 2010-02-03 13:05:46
Description: Yale’s new all-electric building has received the U.S. Green Building Council’s highest rating, the LEED Platinum certification. Kroon Hall, the new home of the Yale School of Forestry & Environmental Studies was designed in a way so that it would use 8
Source: Green Diary

Name: GE Dedicates $45M Renewable Energy Global HQ in NY [score: 8.12215737288]
Publish Date: 2010-02-02 17:23:29
Description: …and Environmental Design (LEED) green building standards and will be 20 percent more energy efficient than required by New York State building standards. Features include low-water faucets, improved insulation, energy-efficient hot water boilers and air conditioning…
Source: Environmental Leader

Name: Bahrain to host Green Building Forum [score: 7.8266672789]
Publish Date: 2010-02-03 16:47:55
Description: …the construction industry and Sustainable building materials; Smart buildings; Renewable energy powering the home; Carbon trading or zero emissions; Innovation in building; Recycling waste; and Green building training and education. The Green Building Forum…
Source: TradeArabia

Name: GE opens Renewable Energy HQ in Schenectady [score: 7.6463159787]
Publish Date: 2010-02-01 20:15:41
Description: After decades of declining numbers at its Schenectady plant, GE is growing new “green” jobs there. GE is turning around a decades-long trend of taking jobs away from its Schenectady plant, opening its $45 million Renewable Energy headquarters in
Source: Fox 23 New York

Query took 2.14235496521 seconds

Results: Segment Transcript #3 (OpenCalais entities)

Name: Israel warns officers after Hamas assassination [score: 10]

Publish Date: 2010-02-01 16:12:38
Description: …during his funeral procession at the Palestinian refugee camp of Yarmouk, near Damascus, Syria, Friday, Jan. 29, 2010. Hamas accused Israeli agents on Friday of assassinating al-Mabhouh, a veteran operative of the Palestinian militant group, saying he was…
Source: KansasCity.com

Name: Hamas-Israel talks on prisoner swap collapse [score: 9.94789268903]
Publish Date: 2010-02-02 22:10:38
Description: …Israeli soldier Gilad Shalit would be traded for about 1,000 of the more than 7,000 Palestinians in Israeli jails. Israel, the Hamas official said, was demanding that dozens of Palestinians imprisoned after being convicted of involvement in lethal attacks…
Source: Lebanon Daily Star

Name: Fatah leader to Hamas-ruled Gaza in goodwill trip [score: 9.6563639293]
Publish Date: 2010-02-03 16:34:38
Description: …force and intentionally” targeted civilians. He said Hamas committed war crimes by firing rockets indiscriminately at Israeli border towns. Hamas insisted in its 80-page response, whose main points were made public last week, that rocket fire was directed…
Source: San Diego Union-Tribune

Name: Israel: Slain Hamas leader key arms smuggler [score: 9.30580435181]
Publish Date: 2010-01-31 15:06:41
Description: …to comment on the allegations against it. Last month, two Hamas men were killed in a mysterious blast in Beirut. Hamas said Israel was a suspect but did not openly accuse it of the killings. The leader of Hamas’ Damascus-based leadership, Khaled Mashaal,…
Source: WHDH-TV

Name: Hamas suspends Shalit talks to protest Dubai assassination [score: 8.99086324777]
Publish Date: 2010-02-03 02:32:33
Description: …been on the brink of an explosion in any case, for which he blamed Israel’s government. Mabhouh, a senior Hamas official who Israel said was involved in smuggling weapons into Gaza, was found dead in a Dubai hotel room on January 20. Hamas has accused Israel…
Source: Haaretz

Query took 0.499907016754 seconds

Results: Segment Transcript #3 (Zemanta entities)

Name: Anticipation mounts at UN over Goldstone report [score: 10]
Publish Date: 2010-02-03 08:26:17
Description: …sense of anticipation within the United Nations (UN) headquarters as the latest chapter of the Goldstone report saga unfolds. Secretary-General Ban Ki-moon is expected to address the General Assembly on Friday on progress made by both Israel and Hamas regarding…

Source: SABC News

Name: Rights groups under fire for scrutiny of Israel’s conduct of Gaza war [score: 9.32449073586]

Publish Date: 2010-02-03 21:07:27
Description: …organizations,” he said. “They are accusing Israel of terrorizing [Palestinian] civilians.” The Goldstone report assigns blame to both Israel and Hamas for committing possible war crimes during the war, but accuses Israel of intentionally killing Palestinian…
Source: AXcess News

Name: Berlusconi: Italy proud of solidarity with Israel [score: 9.10728016456]
Publish Date: 2010-02-03 22:45:26
Description: OCCUPIED JERUSALEM: Premier Silvio Berlusconi Wednesday pledged Italy’s firm support for Israel, urging “effective sanctions” against its arch-foe Iran and speaking out against a damning UN report on the Gaza war. He also called on Israel to halt the
Source: Lebanon Daily Star

Name: Israel: Goldstone report misrepresents our investigative system [score: 8.65828128043]
Publish Date: 2010-01-29 19:23:37
Description: …its own report based on submissions from both sides – following another recommendation from the 547-page Goldstone report that both Israel and Hamas conduct their own investigations. In the report that Israel gave to the UN on Friday, it emphasized that its…
Source: Haaretz

Name: Barak to IDF brass: Goldstone Report biased [score: 8.65329445791]
Publish Date: 2010-02-01 17:15:24
Description: …is biased and false; I unequivocally reject it,” Defense Minister Ehud Barak said Monday just hours after it was reported that two senior IDF officers were disciplined for exceeding their authority during the war in Gaza, a little over a year ago. In face…
Source: Ynetnews

Query took 0.384150981903 seconds

Bing API 2.0 (News Search)

Overview
Currently in pre–release, the 2.0 version of the Bing API allows access to Bing search results for web pages, news, videos and more. There are currently no restrictions on reordering or blending results, and can be used with “customer–facing sites and applications.”

Service Description

Notes

  • Limited query to top 2 entities for this test, due to small result sets (using AND operator)
  • Support for logical operators is inconsistent, though different results are returned for “OR” and “AND” operators.
  • No support for entity weighting
  • No score or relevance value returned with results
  • No URL for source, but can potentially be parsed from article URL

Analysis
The Bing API is still in pre–release (according to the TOU), and feels a bit unpolished. Logical operator use in queries is unclear (and seemingly undocumented), and results are inconsistent. Results, in general, are fairly erratic, as shown below. The lack of any query weighting or filtering capabilities, combined with the fact that Bing reserves the right to insert advertising in the results at any time, excludes this API from use in the ViewChange initiative.

Results: Segment Transcript #1 (OpenCalais entities)

Name: In South Africa, Holidays for FIFA cup with cheap flights to … [score: ]
Publish Date: 2010-02-04T13:57:33Z
Description: PR Log (Press Release) Feb 04, 2010 Johannesburg is major hosting city of FIFA world cup at South Africa. Indicating that the FIFA World Cup in 2010 will be soccer’s better spectacular entertainment. Jo’Burg, Egoli or Johannesburg is a …
Source: PRLog (free press release)

Name: Africa : Australia Upbeat On Doha Pact [score: ]
Publish Date: 2010-02-05T16:26:41Z
Description: Johannesburg — STRONG motivation and political will exist to conclude a broad Doha trade agreement this year, even if some of the details have to be deferred for future negotiation, Australian Trade Minister Simon Crean said yesterday. The Doha …
Source: AllAfrica.com

Name: President’s love child hits a nerve in SAfrica [score: ]

Publish Date: 2010-02-05T17:31:07Z
Description: JOHANNESBURG (AP) — Confirmation that President Jacob Zuma, who has three wives and a fiancee, has fathered a child with yet another woman has prompted jokes in South Africa’s media but has also hit a nerve in a country hardest hit by the virus …
Source: WFAA

Name: Monfils, Ferrer, Lu reach South Africa quarterfinals [score: ]
Publish Date: 2010-02-04T23:23:03Z
Description: JOHANNESBURG — Gael Monfils and David Ferrer continued to plow through the field in reaching the SA Tennis Open quarterfinals on Thursday. Top-seeded Monfils and No. 2 Ferrer have yet to drop a set through two rounds. Monfils beat fellow Frenchman …
Source: CBS Sports

Name: UPDATE 1-S.Africa reserves dip on dollar gains, weaker gold [score: ]
Publish Date: 2010-02-05T06:54:02Z
Description: JOHANNESBURG, Feb 5 (Reuters) – South Africa’s net gold and foreign exchange reserves fell by 0.8 percent to $38.63 billion at the end of January, largely on changes in valuations led by the impact of a stronger dollar. Reserve Bank data on its …
Source: Forex Pros

Query took 0.367772817612 seconds

Results: Segment Transcript #1 (Zemanta entities)

Name: ‘Brick baby’ mom trial to start [score: ]
Publish Date: 2010-02-05T13:13:25Z
Description: Johannesburg – The mother of an 11-month-old baby hit with a brick on the head on New Year’s Eve will go on trial on March 26 in the Hillbrow Magistrate’s Court. Nomalanga Moyo, 28, appeared briefly in court on Friday to face charges of perjury alone …
Source: Beeld

Query took 0.0550220012665 seconds

Results: Segment Transcript #2 (OpenCalais entities)

No results

Results: Segment Transcript #2 (Zemanta entities)

Name: Getting green meetings on the same eco–friendly page [score: ]
Publish Date: 2010–02–05T16:05:13Z
Description: Much as a forest fire clears the land and leaves behind essential nutrients to enrich a new generation of growth, the devastation of the travel and meetings industry caused by a global economic collapse has left a few seedlings. One of them is the …
Source: Green Right Now

Query took 0.133863925934 seconds

Results: Segment Transcript #3 (OpenCalais entities)

Name: Hamas “regrets” civilian deaths, Israel unmoved [score: ]
Publish Date: 2010-02-05T16:41:00Z
Description: GAZA (Reuters) – Hamas, in an unusual move that seems unlikely to herald a change in tactics by the Islamist group, has expressed regret for the deaths of Israeli civilians in Palestinian rocket attacks during fighting in Gaza a year ago. Israel …
Source: Reuters

Name: Israel: Hamas Leader Smuggled Iranian Arms [score: ]
Publish Date: 2010-02-01T13:11:00Z
Description: JERUSALEM — Days after Hamas accused Israel of electrocuting and poisoning one of its commanders in his Dubai hotel room, Israel on Sunday claimed the man played a critical role in smuggling rockets from Iran to Palestinian militants in Gaza …
Source: FOX News

Name: Israel says slain Hamas commander was key figure in smuggling rockets … [score: ]

Publish Date: 2010-01-31T21:04:38Z
Description: JERUSALEM – Days after Hamas accused Israel of electrocuting and poisoning one of its commanders in his Dubai hotel room, Israel claimed Sunday that the dead man played a critical role in smuggling long-range rockets from Iran to Palestinian …
Source: Minneapolis Star Tribune

Name: Hamas threatens to take fight against Israel beyond Gaza [score: ]
Publish Date: 2010-02-02T17:20:21Z
Description: In the wake of a Hamas claim on Friday that Israeli agents assassinated one of its operatives in Dubai last week, the Islamist movement is vowing to take revenge against the Jewish state for the militant’s death even if it means going abroad …
Source: The Christian Science Monitor

Name: Israel, Hamas Respond Differently to Goldstone [score: ]
Publish Date: 2010-02-03T23:17:05Z
Description: Meeting a deadline of February 5 set by the United Nations, Israel and Hamas have offered their responses to what has become known as the Goldstone Report, the controversial 547-page investigation of the Gaza incursion in 2008-09 that accused both …
Source: Forward

Query took 0.389917850494 seconds

Results: Segment Transcript #3 (Zemanta entities)

Name: Goldstone report: Israel and Palestinians respond to UN [score: ]
Publish Date: 2010-01-29T23:01:32Z
Description: Israel and West Bank Palestinians have responded to the UN’s Goldstone report which accused both of them of war crimes during Israel’s Gaza operation. The Israeli defence minister said there was no army as “responsible… moral and accurate…even …

Source: BBC Middle East

Name: UN chief says he can’t determine if Gaza probes by Israel and the … [score: ]

Publish Date: 2010-02-05T17:38:16Z
Description: UNITED NATIONS – U.N. Secretary-General Ban Ki-moon said he could not determine whether the Israelis or Palestinians had conducted credible investigations into allegations of war crimes during last year’s Gaza conflict as required under a U.N …
Source: Minneapolis Star Tribune

Name: Israel,Palestine to respond to Goldstone report [score: ]
Publish Date: 2010-02-05T09:31:31Z
Description: UN Secretary General, Ban Ki-moon UN Secretary General, Ban Ki-moon, says there is not enough evidence yet to say whether Israel and the Palestians are complying with UN demands to investigate the Gaza conflict. They have been asked to respond by …
Source: Ghana Broadcasting Corporation

Name: Israel: Goldstone report mispreresnts our invistgative system [score: ]
Publish Date: 2010-01-29T18:58:09Z
Description: In a response to the Goldstone report, Israel submitted a paper to the UN on Friday that emphasized the inaccuracies of the Goldstone report as well as its misrepresentation of Israel’s investigative system. In a 46 page paper entitled “Gaza …
Source: Haaretz.com

Name: Israel responds to ‘distorted’ Goldstone report – Summary [score: ]

Publish Date: 2010-01-29T15:09:06Z
Description: Jerusalem – Israel Friday called a UN report on its offensive in Gaza last winter “distorted, false and irresponsible,” as it nonetheless met a deadline to report to the international body about probes it had conducted thus far into war crimes …
Source: Earthtimes

Query took 0.341904163361 seconds

Yahoo! BOSS (News Search)

Overview
Yahoo! BOSS (Build your Own Search Service) is an open API for accessing Yahoo’s search platform. Current API methods support searching web pages, news, images and spelling suggestions.

Service Description

Notes

  • Limited query to top 2 entities for this test, due to small result sets
  • Doesn’t seem to support logical OR operator (attempted many query formats, including on the format used by Drupal’s “More Like This” plugin)
  • No score or relevance value returned with results
  • Search narrowing is limited to region, age and language; however, results can be restricted to specific sites
  • Limited documentation & support forum

Analysis
With the BOSS APIs, Yahoo has provided an unlimited white–label search API for developers. While this may prove useful for web searches, the available query options do not provide a very powerful news search API. Again, the lack of query term weighting and effective logical operators severely reduce the potential for relevant news results.

Results: Segment Transcript #1 (OpenCalais entities)

Name: S.Africa-France rugby set for Day 2 of World Cup [score: ]
Publish Date: 2010/02/03 08:00:03
Description: JOHANNESBURG (AP) — With the World Cup coming to Africa for the first time, it was expected to be all football, all the time in South Africa in June and July, but rugby officials have gained permission for a South Africa-France match in Cape Town the day after the World Cup opens.
Source: The News-Times

Name: Ram, Robert advance in Johannesburg [score: ]
Publish Date: 2010/02/02 01:12:07
Description: JOHANNESBURG, South Africa, Feb. 1 (UPI) — Fifth-seeded Rajeev Ram of the United States and No. 8 Stephane Robert of France were among first-round winners Monday in the SA Tennis Open in Johannesburg.
Source: UPI

Name: Fred Bridgland: Wind of change is still blowing across Africa, 50 years on [score: ]
Publish Date: 2010/02/04 00:07:20
Description: THE 50th anniversary of one of the most important speeches in the history of Britain, and also of Africa, passed yesterday with little acknowledgement.
Source: The Scotsman

Name: SOUTH AFRICA: Arms export controls in meltdown [score: ]
Publish Date: 2010/02/02 17:30:09
Description: JOHANNESBURG, 2 February 2010 (IRIN) – A report by South Africa’s Auditor-General has revealed a serious lack of controls over exports of its vast array of conventional weapons, which a non-governmental organisation monitoring the trade attributes to government’s “couldn’t be bothered” attitude.
Source: IRIN

Name: S. Africa Investor Confidence Rises; Fixed-Income Holdings Gain [score: ]
Publish Date: 2010/02/03 09:33:25
Description: Feb. 3 (Bloomberg) — Investor confidence in South Africa rose in the fourth quarter of 2009, with fund managers increasing their fixed-income weightings, according to an index that tracks where investors are moving their money.
Source: Bloomberg

Query took 0.159895181656 seconds

Results: Segment Transcript #1 (Zemanta entities)

No results

Results: Segment Transcript #2 (OpenCalais entities)

Name: And other commentators from [score: ]
Publish Date: 2010/02/04 00:36:01
Description: In reply to Cuba: A Red And Green Utopia? : And other commentators from Cuba have said that most Cubans are, if not happy with constrained existence, then at least going along with it. We really do not need two cars, McMansion houses, electric everything, crammed highways, stinking pollution, and the sooner we wake up to this the better for this Earth. If ever the USA dropped the throttling …
Source: New Matilda

Query took 0.126603126526 seconds

Results: Segment Transcript #2 (Zemanta entities)

Name: Global Warming Is Crap! [score: ]
Publish Date: 2010/01/28 20:31:23
Description: Global warming? says Steve Wampler. Crap! This is unexpected because Steve trained as an environmental engineer at UC Davis. But hes serious.
Source: San Diego Reader

Query took 0.157608985901 seconds

Results: Segment Transcript #3 (OpenCalais entities)

Name: Hamas blames Israel for collapse of talks [score: ]

Publish Date: 2010/02/03 19:59:46
Description: TEL AVIV, Israel, Feb. 3 (UPI) — A senior Hamas official has warned that talks with Israel over a prisoner exchange to free a captured Israeli soldier have collapsed.
Source: UPI

Name: Israel, Hamas Respond Differently to Goldstone [score: ]
Publish Date: 2010/02/03 23:27:41
Description: Meeting a deadline of February 5 set by the United Nations, Israel and Hamas have offered their responses to what has become known as the Goldstone Report, the controversial 547-page investigation of the Gaza incursion in 2008-09 that accused both combatants of war crimes.
Source: Forward

Name: Hamas threatens to take fight against Israel beyond Gaza [score: ]
Publish Date: 2010/02/02 17:08:12
Description: After blaming Israeli agents for assassinating a Hamas official in Dubai on Friday, Hamas said it too could broaden the conflict beyond Israel and Gaza. But analysts are doubtful Hamas can pull it off.
Source: The Christian Science Monitor

Name: Mossad killing of terror chiefs has little impact on Israel-Hamas war [score: ]
Publish Date: 2010/02/03 06:55:32
Description: There is no need for the government of Israel to answer the question of whether Mossad agents were responsible for assassinating Hamas operative Mahmoud al-Mabhouh in Dubai; the smiles on the ministers’ faces as they left the weekly cabinet meeting on Sunday said it all.
Source: Haaretz Daily

Name: Palestinian officials say no pressure to resume talks as Hamas slams PNA-Israel ties [score: ]
Publish Date: 2010/02/03 05:07:44
Description: Palestinian officials said on Tuesday that there is no European pressure on the Palestinian National Authority (PNA) to resume the stalled peace talks with Israel, while Gaza-ruling Islamic Hamas movement slammed ties between the PNA and Israel. Earlier reports said that U.S. peace envoy George Mitchell asked the European Union to exert pressure on Palestinian President Mahmoud Abbas and the PNA …
Source: People’s Daily

Query took 0.207969903946 seconds

Results: Segment Transcript #3 (Zemanta entities)

Name: Alan Dershowitz: The Case Against the Goldstone Report [score: ]

Publish Date: 2010/02/01 17:24:54
Description: The Goldstone Report — commissioned by an organization with a long history of anti-Israel bigotry — is much more scurrilous than most of its detractors (and supporters) believe.
Source: The Huffington Post

Name: Israel to present response to Goldstone report [score: ]
Publish Date: 2010/01/29 15:38:50
Description: The Israeli government will present its response on Friday to the Goldstone report, which suggested Israel may have been guilty of war crimes during its military operation in and around the Gaza Strip a year ago. Many buildings were destroyed and people lost their lives when Israeli tanks and troops moved into the Palestinian coastal enclave in early January, 2009. A week earlier, Israeli …
Source: People’s Daily

Name: Israel, Hamas Respond Differently to Goldstone [score: ]
Publish Date: 2010/02/03 23:27:41
Description: Meeting a deadline of February 5 set by the United Nations, Israel and Hamas have offered their responses to what has become known as the Goldstone Report, the controversial 547-page investigation of the Gaza incursion in 2008-09 that accused both combatants of war crimes.
Source: Forward

Name: Israel writes response to Goldstone report [score: ]
Publish Date: 2010/01/28 20:13:13
Description: Israel has prepared a response to the Goldstone report, which accuses Israel of committing war crimes during the Gaza war.
Source: J Weekly

Name: Berlusconi: Italy is Israel’s big brother [score: ]
Publish Date: 2010/02/03 16:21:25
Description: JERUSALEM (JTA) — The Goldstone report “tried to incriminate Israel for its justified response to Hamas rockets,” Italian Prime Minister Silvio Berlusconi said.
Source: JTA

Query took 0.289761066437 seconds

Content APIs – Videos

Test results are using entities returned from OpenCalais or Zemanta’s entity extraction engine, and have not been curated. In the application environment, these entities will be curated by content managers, so the quality of the results below are not entirely accurate. The number of query terms and the logical operators have been adjusted to return the optimal results for each API. Obviously, there will be much more tuning once the selected API(s) are integrated into the application. The first five video results are shown for each query.

YouTube

Overview
YouTube is the leader in online video, hosting a vast library of user–submitted video content on just about any topic imaginable. The YouTube APIs allow developers to query and incorporate YouTube data into external sites.

Service Description

Notes

  • Limited query to top 3 entities (OR) in the “Nonprofit” category for this test
  • Supports basic logical OR, NOT operators (in both queries and categories), but no weighting
  • No relevance value is returned, however, metrics data such as rating, viewCount, favoriteCount and commentCount are returned
  • Queries can be further limited to specific regions, channels and authors
  • Thumbnail image URLs are returned in both “default” and “hqDefault” sizes

Analysis
The YouTube APIs allow for basic searching of the YouTube video database. While logical operators are supported in queries, the lack of query term weighting can lead to inaccurate results. There are a few options available for narrowing search results, including limiting the search categories, or restricting searches to curated channels or playlists. The available query parameters give the developer some control over the search results, but when you consider the huge spectrum of non–curated user–submitted video, it is difficult to get accurate results from a simple keyword search.

Results: Segment Transcript #1 (OpenCalais entities)

Name: ICOPF 2008 Port Elizabeth South Africa [score: ]
Publish Date: 2010-02-03T05:29:43.000Z
Description: Moments of Blessings Church and International Churches of Prayer Fellowship in Port Elizabeth, South Africa. Pastor Patrick Duncan and MCM Youth Department singing, “I Press” Mount Carmel Ministries International is under the banner of ICOPF one of the many churches Bishop Archie oversees. Pastor Irvan Van Wyk is the pastor and Pastor Desmond Peterson is the Senior Pastor of MCM. Bishop George Archie will be in South Africa to help these pastors in February 2009 to continue the work God has given him. For more info contact: Moments of Blessings Church toll free at 1-866-423-2464.
Source: BishopGeorgeArchie

Name: The Fruits of Fairtrade in South Africa [score: ]
Publish Date: 2010-01-31T15:35:28.000Z
Description: This short film explores how Fairtrade is changing the lives of workers on Fairtrade certified citrus fruit farms in South Africa.
Source: FairtradeMarkIreland

Name: How Not to Write About Africa – Binyavanga Wainaina – narrated by Djimon Hounsou [score: ]

Publish Date: 2009-01-04T19:07:50.000Z
Description: (RED)Wire’s 2nd edition came with this awesome short – here is the blurb. When Bono edited the Africa issue of Vanity Fair, it included an essay written by Kenyan writer Binyavanga Wainaina. Through that, we became aware of another piece he’d written for Granta a number of years ago called “How (Not) to Write About Africa.” Director Jesse Dylan and his company freeform worked with Binyavanga and the Beninois actor Djimon Hounsou to create this filmed performance of the essay. Thanks to W Hotels for the location and Kenyan musician Ayub Ogada for the music. Read the entire essay at www.granta.com
Source: quaerentia

Name: Jennifer Connelly in charity: water Clean Water Africa PSA [score: ]
Publish Date: 2008-04-04T16:55:08.000Z
Description: imagine if New York City’s taps went dry. What would we do? Jennifer Connelly walks to Central Park to get dirty water for her family as millions of mothers in Africa do every day. This new PSA from charity: water was directed by Hotel Rwanda’s Terry George, cinematography by Ellen Kuras. Music by Rumor Mill. Edited by Michael Rothman @ Mudbutter. Produced by www.publicaddress.tv All involved DONATED their time. Its national commercial debut was on American Idol Gives Back on April 9. Want to act? Only $20 can give one person clean and safe drinking water for 20 years. charity: water helps build wells in Africa and provides clean, safe drinking water. please help us dig wells. Start by helping one person. Find out more at www.charitywater.org
Source: charitywater

Name: A Fight Against Poverty in Africa [score: ]
Publish Date: 2008-02-27T04:26:59.000Z
Description: South Africa has two extremes the good and the bad I may have only experienced one of them but heck… I’ve seen em both Almost 2.5 million people have joined the fight against poverty. Now you have the opportunity to make poverty history. www . one . org you can make a difference ***This video only illustrates one part of Africa. It illustrates the part of Africa that is slowly dying because of poverty and disease. Though much of Africa has been struck by poverty, many areas in Africa are wealthy and economically stable. In fact, some parts of Africa greatly resemble the wealthier portions of the United States.
Source: Scarrface32

Query took 0.255810976028 seconds

Results: Segment Transcript #1 (Zemanta entities)

Name: IDP portion of the Africa trip [score: ]
Publish Date: 2010-01-30T22:26:06.000Z
Description:
Source: asharifzadeh6

Name: Africa Trip [score: ]
Publish Date: 2010-02-01T03:52:04.000Z
Description: Our trip to Malawi, Africa with Children of the Nations
Source: mpayne1970

Name: How Not to Write About Africa – Binyavanga Wainaina – narrated by Djimon Hounsou [score: ]
Publish Date: 2009-01-04T19:07:50.000Z
Description: (RED)Wire’s 2nd edition came with this awesome short – here is the blurb. When Bono edited the Africa issue of Vanity Fair, it included an essay written by Kenyan writer Binyavanga Wainaina. Through that, we became aware of another piece he’d written for Granta a number of years ago called “How (Not) to Write About Africa.” Director Jesse Dylan and his company freeform worked with Binyavanga and the Beninois actor Djimon Hounsou to create this filmed performance of the essay. Thanks to W Hotels for the location and Kenyan musician Ayub Ogada for the music. Read the entire essay at www.granta.com
Source: quaerentia

Name: A Fight Against Poverty in Africa [score: ]
Publish Date: 2008-02-27T04:26:59.000Z
Description: South Africa has two extremes the good and the bad I may have only experienced one of them but heck… I’ve seen em both Almost 2.5 million people have joined the fight against poverty. Now you have the opportunity to make poverty history. www . one . org you can make a difference ***This video only illustrates one part of Africa. It illustrates the part of Africa that is slowly dying because of poverty and disease. Though much of Africa has been struck by poverty, many areas in Africa are wealthy and economically stable. In fact, some parts of Africa greatly resemble the wealthier portions of the United States.
Source: Scarrface32

Name: Making Bamboo Bikes in Africa [score: ]
Publish Date: 2008-04-07T06:22:32.000Z
Description: Craig Calfee introduces bamboo bikes to Ghanaians. Cargo bikes are built and test ridden with heavy loads. A bamboo reinforced wheel is also demonstrated.
Source: craigcalfee

Query took 0.227313995361 seconds

Results: Segment Transcript #2 (OpenCalais entities)

Name: Bringing Electricity to Afghanistan – Juan Miranda, ADB [score: ]
Publish Date: 2009-08-21T00:02:08.000Z
Description: Juan Miranda, Director General of of adb’s Central and West Asia Department, talks about adb’s role in bringing electricity to Afghanistan.
Source: AsianDevelopmentBank

Name: Electricity from biomass [score: ]
Publish Date: 2008-01-14T12:26:35.000Z
Description: Electricity from biomass in Bihar, India
Source: myclimate

Name: Energy is Life: Electricity Changing Lives in Afghanistan – ADB (photo essay) [score: ]
Publish Date: 2009-08-24T04:12:32.000Z
Description: In January 2009, electricity began to flow into Kabul along a newly constructed transmission line running from neighboring Uzbekistan. For the first time in more than a generation, many of the capitals 4 million people can now enjoy the benefits of power. Helping to bring electricity to the people of Afghanistan is a key component of adbs strategy to support Afghanistans reconstruction and development.
Source: AsianDevelopmentBank

Name: The Elements – Episode 3: The Ultimate Electricity Absorbing Apparatus [score: ]
Publish Date: 2009-11-30T06:30:14.000Z
Description: Mr. Voltage is sent out by Dr. Ego to fullfil his electricity overconsumption task. Will he succeed? Watch the 3rd episode of the Elements!
Source: homeoftheelements

Name: To Our Leaders: Give Us 100% Clean Electricity in 10 Years [score: ]
Publish Date: 2008-08-15T17:21:43.000Z
Description: We must save our economy, lower fuel costs, free ourselves from our addiction to oil, and solve the climate crisis. To do this, we must demand that we Repower America with 100% clean electricity within 10 years. Go to www.wecansolveit.org Song: Affirmation Composer: De Wolfe Music
Source: WeCanSolveIt

Query took 0.307240009308 seconds

Results: Segment Transcript #2 (Zemanta entities)

Name: Grocery Store Wars (2005) [score: ]
Publish Date: 2006-11-14T21:46:24.000Z
Description: Not long ago in a supermarket not so far away. Help fight the dark side of the farm. Rate the film, favorite the film, comment the film and subscribe to our channel for the freshest Free Range films.
Source: FreeRangeStudios

Name: UN Foundation Green Building [score: ]
Publish Date: 2008-02-05T22:31:38.000Z
Description: UN Foundation has moved into a Gold standard LEED-certified office in early 2007, which will further reduce the Foundation’s impact on the environment while maximizing efficiency, savings and staff comfort.
Source: ThePeopleSpeak

Name: Renewable energy at Manchester City [score: ]
Publish Date: 2008-02-18T11:53:21.000Z
Description: Premiership football club Manchester City will soon have the world’s first stadium powered by renewable energy. They’re in the process of installing a 120m high 3MW turbine which will meet all their energy needs and allow them to sell electricity back to the grid.
Source: GreenpeaceUK

Name: Eco Biz – SB GREEN BUILDING [score: ]
Publish Date: 2008-05-07T21:08:23.000Z
Description: www.sundancechannel.com Joe Sprouls gives a tour of Citigoup’s first Green Sky Scraper. ECOBIZ airs Tuesdays at 9PM on Sundance Channel during “Big Ideas For A Small Planet”.
Source: sundancechannel

Name: Development Marketplace – Uganda: Renewable Energy-Powered Milk Coolers [score: ]
Publish Date: 2008-10-01T18:31:02.000Z
Description: William Kisaalita of the University of Georgia discusses his project to test a new milk cooling system with a pilot group of 50 small-scale farmers to keep milk fresh overnight so it can be safely transported
Source: WorldBank

Query took 0.307975053787 seconds

Results: Segment Transcript #3 (OpenCalais entities)

Name: Birthright Israel NEXT Austin Training 2009 [score: ]
Publish Date: 2010-02-01T23:57:16.000Z
Description:
Source: birthrightisrael

Name: Obama Defends His Backing Of Israel And Thier Terrorist War Against Palestine – He Is A Disgrace [score: ]
Publish Date: 2010-02-01T22:13:05.000Z
Description: Here we have the teleprompter presidents strugging to answer a question from a student. “The middle east is a problem that has plagued the region for centuries” – what in hells name is he on about? im sure he doesnt know either. He should be tried in the hague along with all the other war mongering bastards that think they are above us and the law.
Source: celtickev999

Name: Chinese Jews from Kaifeng arrive in Israel 2009 – a moving documentary [score: ]
Publish Date: 2009-11-04T16:33:07.000Z
Description:
Source: ShaveiIsrael

Name: In a democracy, we’re all parts of the same body [score: ]
Publish Date: 2009-01-31T17:18:39.000Z
Description: A video about democracy. Country: Brazil Contestant: Anna Carolina dos Santos Israel E-mail: annacarol@mail.com
Source: ACSIsrael

Name: Israel’s “Colonies” – Anna Baltzer [score: ]
Publish Date: 2008-10-21T16:43:58.000Z
Description: Extracts of A Witness in Palestine, written by Anna Baltzer. Anna Baltzer, a young Jewish American, went to the West Bank to discover the realities of daily life for Palestinians under the occupation. What she found would change her outlook on the conflict forever. She wrote this book to give voice to the stories of the people who welcomed her with open arms as their lives crumbled around them. For five months, Baltzer lived and worked with farmers, Palestinian and Israeli activists, and the families of political prisoners, traveling with them across endless checkpoints and roadblocks to reach hospitals, universities, and olive groves. Baltzer witnessed firsthand the environmental devastation brought on by expanding settlements and outposts and the destruction wrought by Israels Security Fence, which separates many families from each other, their communities, their land, and basic human services. What emerges from Baltzers journal is not a sensationalist tale of suicide bombers and conspiracies, but a compelling and inspiring description of the trials of daily life under the occupation. Anna Baltzer is a Jewish American graduate of Columbia University, Fulbright scholar, and two-time volunteer with the International Womens Peace Service in the West Bank, where she documented human rights abuses and supported the nonviolent resistance movement to the occupation. Check the playlist I have shown herein in my channel. It’s the full video, almost. /|\
Source: redrose222

Query took 0.263328075409 seconds

Results: Segment Transcript #3 (Zemanta entities)

Name: Birthright Israel NEXT Austin Training 2009 [score: ]
Publish Date: 2010-02-01T23:57:16.000Z
Description:
Source: birthrightisrael

Name: Obama Defends His Backing Of Israel And Thier Terrorist War Against Palestine – He Is A Disgrace [score: ]
Publish Date: 2010-02-01T22:13:05.000Z
Description: Here we have the teleprompter presidents strugging to answer a question from a student. “The middle east is a problem that has plagued the region for centuries” – what in hells name is he on about? im sure he doesnt know either. He should be tried in the hague along with all the other war mongering bastards that think they are above us and the law.
Source: celtickev999

Name: Chinese Jews from Kaifeng arrive in Israel 2009 – a moving documentary [score: ]
Publish Date: 2009-11-04T16:33:07.000Z
Description:
Source: ShaveiIsrael

Name: The UN Blood Libel: The Goldstone Report at the UN Human Rights Council [score: ]
Publish Date: 2009-10-07T13:21:18.000Z
Description: September 29, 2009 Geneva UN Human Rights Council
Source: eyeontheun

Name: In a democracy, we’re all parts of the same body [score: ]
Publish Date: 2009-01-31T17:18:39.000Z
Description: A video about democracy. Country: Brazil Contestant: Anna Carolina dos Santos Israel E-mail: annacarol@mail.com
Source: ACSIsrael

Query took 0.34356713295 seconds

Truveo

Overview
Truveo is a video search engine that indexes “over 300 million videos from thousands of sources across the web.” An AOL subsidiary, Truveo powers video search for AOL Video, Brightcove, Microsoft Corporation, CNET and many smaller sites. According to CrunchBase, Truveo has a sharper focus on professional content, prioritizing commercially produced media over user–generated content. Several mashup applications have been created by independent developers using Truveo’s open APIs.

Service Description

Notes

  • Limited query to top 3 entities (OR) for this test
  • Supports logical operators in search queries: AND, OR, NOT — OR operators are evaluated before AND operators
  • Search queries can be narrowed by category, channel (source), title or any available metadata field
  • Search results can be filtered by video type (free, pay), age, site, etc. and sorted by several criteria including popularity and recency
  • Video URL result is a redirection through http://xml.truveo.com, not parseable
  • Bonus: returns Link TV video results

Analysis
Though the Truveo API does not support query term weighting, it does provide a rich set of filtering and sorting options. Truveo pulls from “professional” media outlets like CNN and WSJ, as well as user–based content from YouTube (and potentially others), and provides a more full–featured query API than YouTube or the other video APIs mentioned here.

Results: Segment Transcript #1 (OpenCalais entities)

Name: World Cup Countdown With Ruud Gullit [score: 385]
Publish Date: January 27, 2010 1:12:08 PM PST GMT
Description: ESPN soccer analyst Ruud Gullit talks about the World Cup coming to South Africa and the chances of the American team getting out of the first round
Source: ESPN

Name: South Africans give up guns [score: 359]
Publish Date: February 5, 2010
Description: South Africa hopes to reduce the amount of guns on the streets in a country where 50 people are murdered daily.
Source: CNN

Name: The 411 on The Grammy Winners, Invictus Premiere, John Terry and more – Mon 1st Feb [score: 385]
Publish Date: 2010-02-01T17:31:29.000Z
Description: Round-up of today’s showbiz news including: Invictus, John Terry and more – Mon 1st Feb. Follow us on twitter at http://twitter.com/theshowbiz411
Source: YouTube

Name: Zak Feau’nati on playing Jonah Lomu in Invictus [score: 385]
Publish Date: 2010-02-01T16:29:15.000Z
Description: The Samoan rugby player talks about his role as Jonah Lomu in Invictus, a film about the 1995 World Cup in South Africa. . Follow us on twitter at twitter.com
Source: YouTube

Name: Davos: Jacob Zuma Promotes World Cup at Forum [score: 388]
Publish Date: Fri, 29 Jan 2010 17:00:44 GMT
Description: South African President Jacob Zuma goes on a public relations blitz at the Davos economic forum to promote the World Cup in South Africa while. WSJ reporter Roman Kessler has more.
Source: The Wall Street Journal

Query took 0.345711946487 seconds

Results: Segment Transcript #1 (Zemanta entities)

Name: Lloyd Banks Gets Air Sick Over Africa [2008] [score: 306]
Publish Date:
Description:
Source: MTVM

Name: De-Mining Southern Sudan [score: 332]
Publish Date: February 3, 2010
Description: David McKenzie has a special look into Southern Sudan and takes us into the world of mine clearance.
Source: CNN

Name: On the Road to Healing [score: 332]
Publish Date: February 3, 2010
Description: David McKenzie explains the idea that could drive Sudanese health care forward.
Source: CNN

Name: Madonna and Malawi [score: 326]
Publish Date: January 26, 2010
Description: Alina Cho talks to Madonna about her efforts to help orphans in Malawi.
Source: CNN

Name: South Africans give up guns [score: 315]
Publish Date: February 5, 2010
Description: South Africa hopes to reduce the amount of guns on the streets in a country where 50 people are murdered daily.
Source: CNN

Query took 0.0722270011902 seconds

Results: Segment Transcript #2 (OpenCalais entities)

Name: Lights still out in Haiti [score: 304]
Publish Date: January 28, 2010
Description: CNN’s Ivan Watson reports on efforts to return electricity to Port-au-Prince, Haiti.
Source: CNN

Name: Arctic Blast Hits Midwest [score: 303]
Publish Date: January 29, 2010 4:27 PM
Description: Across Oklahoma, at least 179,000 homes have no electricity and ice-slicked roads have caused hundreds of accidents. Don Teague reports.
Source: CBS News

Name: Ratcliffe Says Southern May Get 70% in Loan Guarantees: Video [score: 320]
Publish Date: 2010-02-03T00:21:14.000Z
Description: Feb. 2 (Bloomberg) — David Ratcliffe, chief executive officer at Southern Co., talks with Bloomberg’s Carol Massar and Matt Miller about President Barack Obama proposal for additional nuclear-power loan guarantees in his 2011 budget. Ratcliffe said Southern expects to get federal loan guarantees worth as much as $3.5 billion. (Source: Bloomberg)
Source: YouTube

Name: Chu Says Budget May Fund 7 to 10 New Nuclear Reactors: Video [score: 320]
Publish Date: 2010-02-01T20:56:41.000Z
Description: Feb. 1 (Bloomberg) — US Energy Secretary Steven Chu talks with Bloomberg’s Peter Cook about President Obama’s proposal for $54 billion in nuclear-power plant loan guarantees. Chu also discusses Obama’s plan to end tax credits and deductions for domestic oil and gas production and the prospects for natural gas. (Source: Bloomberg)
Source: YouTube

Name: Long Island University Extension – Charlie Ciuffo (Evolution) // video added February 02, 2010 // 0 comments // // Embed video: [score: 320]
Publish Date:
Description: Evolution of innovation correlation
Source: current tv

Query took 0.0815749168396 seconds

Results: Segment Transcript #2 (Zemanta entities)

Name: How to grow more renewable energy [score: 343]
Publish Date: Fri, 05 Feb 2010 11:34:58 -0500
Description: Feb 5 – Researchers are trying to breed plants that could be better sources of renewable energy. A team at Aberystwyth University in Wales is looking improving yields of fast growing plants without increasing inputs such as fertilizers.
Source: Reuters

Name: Citizen Engineer [score: 402]
Publish Date: 2010-02-05T23:25:25.000Z
Description: (October 14, 2009) “Whereas the 20th century belonged to the scientist, the 21st century”, says Sun Microsystems CTO Greg Papadopoulos, “is the domain of the engineer.” Rather than secretly toiling away on new discoveries, modern engineers are concerned about social responsibility, renewable materials and product life cycles, collaborative and open source discovery, and furthering industry-wide innovation. As Chief Technology Officer and Executive Vice President of Research and Development at Sun, Greg Papadopoulos directs the company’s approximate $2B in R&D portfolio with an eye toward innovation, simplicity and eco-responsibility. Stanford University: www.stanford.edu Stanford Engineering Everywhere: see.stanford.edu Stanford Center for Professional Development: scpd.stanford.edu Stanford University Channel on youtube: www.youtube.com
Source: YouTube

Name: Doors and Windows [score: 392]
Publish Date: 2009-04-02T20:30:47Z
Description: The most responsible way to build using wood, a dwindling natural resource, is to not use too much of it. Fortunately, because Kevin is building a straw bale home.
Source: Hulu

Name: Lighting [score: 394]
Publish Date: 2009-04-02T20:30:46Z
Description: Building Green health expert Alyssa Alvord explains that electromagnetic frequency is all around us and discusses the dangers of low-level exposure.
Source: Hulu

Name: Framing and Roofing [score: 392]
Publish Date: 2009-04-02T20:30:47Z
Description: Kevin visits green architect Eric Corey Freed to discuss the environmental differences between using wood and steel for the post-and-beam structure of his new home.
Source: Hulu

Query took 0.0922930240631 seconds

Results: Segment Transcript #3 (OpenCalais entities)

Name: Mosaic News – 1/29/10: World News From The Middle East [score: 380]
Publish Date:
Description: Mosaic is a Peabody Award-winning daily news show covering the Middle East. Tonight’s top stories: Bin Laden the environmentalist- Dubai ‘identifies suspects in Hamas commander killing’- African Nations Cup: Egypt win over Algeria sends national pride soaring.Bin Laden the environmentalistAl Jazeera TV, QatarBritain to stop exporting useless bomb detectors to IraqBaghdad TV, IraqUS Congress should not interfere in Arab mediaAl Aqsa, GazaDubai ‘identifies suspects in Hamas commander killing’ Al Arabiya TV, UAEAfrican Nations Cup: Egypt win over Algeria sends national pride soaring BBC- ArabicHaiti: AView from the Middle EastLink TV, USA
Source: Link TV

Name: Diameter of the Bomb [score: 306]
Publish Date: 2010-02-05T20:31:07Z
Description: This is the story of the bombing of bus 32 in Jerusalem in June 2002. The film connects the stories of a group of ordinary Israelis-Jews and Arabs. Each of them holds a clue to someone who died that day.
Source: Hulu

Name: Story Hole: Children’s Cartoons from Hamas [score: 346]
Publish Date:
Description: Hamas might be making antiSemitic cartoons as payback for Dr Bagelmans old kids show Jewby Doo
Source: IMDb

Name: Mosaic News – 2/4/10: World News From The Middle East [score: 341]
Publish Date:
Description: Mosaic is a Peabody Award-winning daily compilation of television news reports from the Middle East, including Egypt, Lebanon, Israel, Syria, the Palestinian Authority, Iraq and Iran.Sha’ath hopes Gaza trip will set Hamas-Fatah reconciliation? Dubai TV, UAEIsrael’s Lieberman cautions SyriaJordan TV, JordanAssassination in Dubai will only make Hamas strongerSyria TV, SyriaCarbon monoxide poisoning in JordanDubai TV, UAENATO to launch offensive against TalibanAl Jazeera TV, QatarIran: US fighting “psychological war” in GulfPress TV, IranMauritania: progress with al-Qaeda prisonersAl Arabiya TV, UAEFrance denies citizenship to man with veiled wifeBBC- Arabic’Ajami’ nominated for Oscar IBA TV, Israel
Source: Link TV

Name: Israel Weighs Options in Iran Nuclear Threat [score: 365]
Publish Date: Fri, 05 Feb 2010 23:49:21 GMT
Description: For the last few years, Israel has made the case that Iran poses a serious threat to its existence and it has not ruled out the possibility of a preemptive strike against Iran’s nuclear facilities. Tehran denies it is enriching uranium for weapons and says it will retaliate if attacked. The failure of the major powers to engage Teheran and convince it to stop developing its atomic capabilities is raising anxiety among Israeli leaders who are trying to determine whether to strike or give diplomacy another chance. As VOA Jerusalem correspondent Luis Ramirez reports, Israel is drawing on lessons from its past as it deals with what may be a new threat in the not-too-distant future.
Source: voa

Query took 0.0776679515839 seconds

Results: Segment Transcript #3 (Zemanta entities)

Name: Diameter of the Bomb [score: 305]
Publish Date: 2010-02-05T20:31:07Z
Description: This is the story of the bombing of bus 32 in Jerusalem in June 2002. The film connects the stories of a group of ordinary Israelis-Jews and Arabs. Each of them holds a clue to someone who died that day.
Source: Hulu

Name: Mosaic News – 1/29/10: World News From The Middle East [score: 345]
Publish Date:
Description: Mosaic is a Peabody Award-winning daily news show covering the Middle East. Tonight’s top stories: Bin Laden the environmentalist- Dubai ‘identifies suspects in Hamas commander killing’- African Nations Cup: Egypt win over Algeria sends national pride soaring.Bin Laden the environmentalistAl Jazeera TV, QatarBritain to stop exporting useless bomb detectors to IraqBaghdad TV, IraqUS Congress should not interfere in Arab mediaAl Aqsa, GazaDubai ‘identifies suspects in Hamas commander killing’ Al Arabiya TV, UAEAfrican Nations Cup: Egypt win over Algeria sends national pride soaring BBC- ArabicHaiti: AView from the Middle EastLink TV, USA
Source: Link TV

Name: Mosaic News – 1/28/10: World News From The Middle East [score: 360]
Publish Date:
Description: Mosaic is a Peabody Award-winning daily compilation of television news reports from the Middle East, including Egypt, Lebanon, Israel, Syria, the Palestinian Authority, Iraq and Iran.Karzai reaches out to TalibanAl Arabiya TV, UAEPakistan seeks role as mediator with TalibanAl Arabiya TV, UAENATO chief hails results of London conference on Afghanistan Press TV, IranIsrael prepares answer to Goldstone ReportAl Jazeera TV, QatarIran executes two over post election protestsBBC- ArabicObama reiterates need for more jobs, stronger economy Jordan TV, JordanDarfur peace talks resumed in DohaSudan TV, SudanImam kidnapped in LebanonNew TV, LebanonBirds: a profitable trade in PakistanDubai TV, UAE
Source: Link TV

Name: Press TV- News Analysis – Israeli Role in Middle East -03-02-2010 (Part 1) [score: 426]
Publish Date: 2010-02-05T13:44:52.000Z
Description: Press, TV-News, Analysis-Middle, East-Israel-IDF-War, Crimes-Gaza, War-Occupied, Territory-Media, Prejudice-Goldstone, Report
Source: YouTube

Name: Israelis back new strike on Gaza [score: 426]
Publish Date: 2010-02-03T17:38:48.000Z
Description: A recent poll has revealed more than half of Israelis approve of another attack on Gaza. That’s despite widespread international condemnation of last year’s offensive which killed more than 1300 Palestinians. It comes as recent Goldstone Gaza report said Israel committed war crimes, and there’ve been many claims of the use of banned weapons.
Source: YouTube

Query took 0.182127952576 seconds

Bing API 2.0 (Video Search)

Overview
See description above

Service Description
See description above

Notes

  • Limited query to top 2 entities for this test (AND)
  • Support for logical operators is inconsistent, though different results are returned for “OR” and “AND” operators.
  • No support for entity weighting
  • No score or relevance value returned with results
  • No URL for source, but can be parsed from video URL
  • No date or description text provided with results
  • SourceTitle only comes through for large sources (ie. YouTube, Dailymotion), not smaller sites
  • Thumbnail JPEG image provided at 160×120

Analysis
As with the Bing News Search results (same API), the results from the Bing Video Search API are inconsistent. The video results contain less metadata than the news API results, lacking date, description text and source details.

Results: Segment Transcript #1 (OpenCalais entities)

Name: Johannesburg – South Africa, Rio de Janeiro – Brazil, Warnemunde … [score: ]
Publish Date:
Description:
Source: Dailymotion

Name: Ethiopian in South Africa Johannesburg praise and worship #1 [score: ]
Publish Date:
Description:
Source: YouTube

Name: Johannesburg – South Africa, Auckland – New Zealand, Saint-Malo … [score: ]
Publish Date:
Description:
Source: Dailymotion

Name: Johannesburg Takeoff ; Johannesburg from air [score: ]
Publish Date:
Description:
Source: YouTube

Name: Johannesburg – Boksburg – South Africa [score: ]
Publish Date:
Description:
Source: YouTube

Query took 0.310290813446 seconds

Results: Segment Transcript #1 (Zemanta entities)

Name: Epilogue to Hillbrow RIP [score: ]
Publish Date:
Description:
Source: YouTube

Name: The New South Africa [score: ]
Publish Date:
Description:
Source: YouTube

Name: Sunrise in Johannesburg [score: ]
Publish Date:
Description:
Source: YouTube

Name: Receive Me Hillbrow – Part 1 of 2 [score: ]
Publish Date:
Description:
Source: YouTube

Name: Receive Me Hillbrow – Part 2 of 2 [score: ]
Publish Date:
Description:
Source: YouTube

Query took 0.149505853653 seconds

Results: Segment Transcript #2 (OpenCalais entities)

Name: Our Most Ambitious Project [score: ]
Publish Date:
Description:
Source: MySpace

Name: The Lazy F Ranch [score: ]
Publish Date:
Description:
Source: YouTube

Query took 0.0688650608063 seconds

Results: Segment Transcript #2 (Zemanta entities)

Name: TakePart Social Issues: Dolphin, Whale & Porpoise Hunting, Facts on … [score: ]
Publish Date:
Description:
Source:

Name: THE REAL G! – PodcastBlaster Podcast Directory [score: ]
Publish Date:
Description:
Source:

Name: THE REAL G! – PodcastBlaster Podcast Directory [score: ]
Publish Date:
Description:
Source:

Name: THE REAL G! – PodcastBlaster Podcast Directory [score: ]
Publish Date:
Description:
Source:

Name: THE REAL G! – PodcastBlaster Podcast Directory [score: ]
Publish Date:
Description:
Source:

Query took 0.0580070018768 seconds

Results: Segment Transcript #3 (OpenCalais entities)

Name: The beauty of Hamas and the ugliness of Israel 1/2 [score: ]
Publish Date:
Description:
Source: Dailymotion

Name: Hamas’ Israeli Hostage [score: ]
Publish Date:
Description:
Source: CBS News

Name: Israel May Invade Gaza [score: ]
Publish Date:
Description:
Source: CBS News

Name: Israel Hits 1,000 Hamas” targets [score: ]
Publish Date:
Description:
Source: ABC News

Name: Israel and Hamas vow to fight on in Gaza [score: ]
Publish Date:
Description:
Source: Dailymotion

Query took 0.484480142593 seconds

Results: Segment Transcript #3 (Zemanta entities)

Name: Peres favors diplomacy [score: ]
Publish Date:
Description:
Source: CNN

Name: Anger at Abbas over Goldstone report delay – 05 Oct 09 [score: ]
Publish Date:
Description:
Source: YouTube

Name: Richard Falk on Palestine and Goldstone report – 07 Oct 09 [score: ]
Publish Date:
Description:
Source: YouTube

Name: UN Goldstone report on Gaza predictably biased vs Israel Jewu 512 [score: ]
Publish Date:
Description:
Source: YouTube

Name: Banned Speech: The UN Council That Created the Goldstone Report [score: ]
Publish Date:
Description:
Source: YouTube

Query took 0.0801150798798 seconds

Vimeo

Overview
Vimeo is a video sharing site for user–submitted content. With almost 3 million members, and a high volume of new video content uploaded daily, it is a widely–adopted video platform, popular with blog owners. There seems to be more of a community focus than competing sites, such as YouTube.

Service Description

Notes

  • Limited query to top 2 entities (AND) for this test
  • Support for logical operators is unclear, though different results are returned for “OR” and “AND” operators
  • Somewhat cumbersome oAuth authentication for simple search query
  • No score or relevance value returned with results
  • Publish date shown in results is the upload date
  • Limited documentation

Analysis
The videos.search method of the Vimeo Advanced API is fairly limited in terms of query building. There are no (documented) parameters for narrowing the search results, or weighting search terms. The results below are probably the least relevant of all the APIs tested, likely because of the randomness associated with user–submitted videos, which makes up the entire search pool. With limited query parameter support, and a library of entirely user–submitted video, the probability of returning relevant video results is extremely low.

Results: Segment Transcript #1 (OpenCalais entities)

Name: REHEARSAL LAST SUPPER JOHANNESBURG 2007 [score: ]
Publish Date: 2009-09-26 06:18:27
Description: REHEARSAL LAST SUPPER was produced in Johannesburg in march 2007 and presented as an art video installation in the Artist-In-Residence space ‘The Bagfactory’. Young actors from the Market Theatre Laboratory and the Witwatersrand University Drama Department participated in this project, an improvisat
Source: ANKE SCHAEFER (Vimeo)

Name: HCCI Clubhouse Shout Out to Johannesburg, South Africa [score: ]
Publish Date: 2008-08-20 12:03:09
Description: When a new computer clubhouse came online in Johannesburg a few years ago, the HCCI Computer Clubhouse was selected to take part in a special presentation video for their ribbon cutting ceremonies. This video from our Clubhouse members welcoming the new Clubhouse to the network was produced by Inte
Source: HCCI Computer Clubhouse (Vimeo)

Name: Paul petting a turtle and a goat in Gold Reef City, Johannesburg, South Africa [score: ]
Publish Date: 2008-12-14 12:51:57
Description: Stroking a turtle, when a goat comes up and says hello!
Source: Paul Richardson (Vimeo)

Name: Jobusy Adventure #4 The Africa Day Concert [score: ]
Publish Date: 2009-05-25 12:48:16
Description: Held every year on the 25th of May, Africa Day aims to help celebrate the diversity of african cultures. The team joined the masses in Mary Fitzgerald Square in Newtown Johannesburg, and partied all night to the best jams coming out of the continent.
Source: russell grant (Vimeo)

Name: Goat and Cat say Hello [score: ]
Publish Date: 2008-12-21 11:59:00
Description: A goat and a cat say hello to each other in Gold Reef City, Johannesburg, South Africa
Source: Paul Richardson (Vimeo)

Query took 0.830136060715 seconds

Results: Segment Transcript #1 (Zemanta entities)

Name: DREAMBOY [score: ]
Publish Date: 2009-11-16 09:11:33
Description: During a theatre production in South Africa, Geert Mul had handed out ten digital video cameras to street-kids of Hillbrow, Johannesburg. He asked them to portray their friends and their surroundings. Out of the 14 hours of dramatic video material that returned, Mul selected seconds to create the vi
Source: Geert Mul (Vimeo)

Query took 0.315944910049 seconds

Results: Segment Transcript #2 (OpenCalais entities)

Name: Feel the Beat! – The Vibe of EPIC ‘08 [score: ]
Publish Date: 2008-04-28 13:38:54
Description: It’s a consumer show of EPiC proportions. A bad pun (I know, I know) but I have never been to a show like this, and honestly I was blown away. I missed EPIC last year, so I wasn’t sure what to expect when I walked through the crimson arches yesterday. The atmosphere is warm and inviting. The vibe
Source: Happyfrog.ca (Vimeo)

Name: City Attorney Clashes with Speaker at 2009 Congress of Neighborhoods -Accused of Bullying, Questioned on Red Ant Theory [score: ]
Publish Date: 2009-10-12 21:01:48
Description: [Links to sources and videos at end of article ] Los Angeles, City Attorney, Carmen Trutanich was taking questions from the floor Oct 10, 2009 at the 2009 Congress of Neighborhoods held at City Hall in Los Angeles, California. Trutanich and a woman, later identifying herself as Kathryn Schorr a
Source: Michael N Cohen (Vimeo)

Name: Burning Man: 2002 Mega-volt from the film AquaBurn by Bill Breithaupt [score: ]
Publish Date: 2008-12-10 05:06:06
Description: Burning Man Festival 2002: This segment is from the Burning Man film “AQUABURN”,directed by Bill Breithaupt. The segment is about the crazy world of Dr. MegaVolt. Burningman’s most beloved ICON . AquaBurn is an award-winning documentary film by director Bill Breithaupt showcasing “The Floating Wor
Source: Bill Breithaupt (Vimeo)

Name: vimeo vimeo [score: ]
Publish Date: 2009-12-23 23:56:43
Description: White people can do powerful things with their eyes: casting judgment, indicating scorn, and obnoxiously rolling them when someone says something they don’t agree with. Yet in spite of these powers, they are not immune to the dangers of the sun. So white people must wear sunglasses. But what may s
Source: bluenote736 (Vimeo)

Query took 1.13659310341 seconds

Results: Segment Transcript #2 (Zemanta entities)

Name: Eric Corey Freed of Organic Architect On the Future of Green Building [score: ]
Publish Date: 2009-12-03 00:36:31
Description: Did you know that buildings have more impact on the environment than cars? At West Coast Green, Eric Corey Freed of Organic Architect talks about the future of green building and how it is possible to have a completely nontoxic, organic home. Eric Corey Freed (ECF): You know nearly half of carb
Source: Lorna Li (Vimeo)

Name: Cottages on Greene [score: ]
Publish Date: 2009-10-13 10:02:07
Description: Nestled in the Historic Hill and Harbor District of East Greenwich, 15 units of mixed-income condominiums have been organized into a compact cottage court development in a variety of building type on a vacant lot. Small, efficient and well detailed, this development, replete with rain gardens to in
Source: Matthew Valero, LEED AP (Vimeo)

Name: GO >> Embark on The Journey [score: ]
Publish Date: 2009-07-02 18:09:28
Description: The Marcus Graham Project cordially invites you to embark on an impromptu journey of knowledge, ingenuity, creativity & self expression in an event entitled gO Friday, July 10th 2009 7p 11p The South Side on Lamar 1409 S. Lamar Loft 111 Please join us for a night of Green Edu-tainment a
Source: Marcus Graham (Vimeo)

Name: Summer Garden Foods [score: ]
Publish Date: 2007-08-25 06:39:52
Description: Published: Tuesday, August 21, 2007 Food manufacturer to expand The six-acre site on McClurg Road will house the Summer Garden corporate offices and a processing plant. By WILLIAM K. ALCORN VINDICATOR STAFF WRITER BOARDMAN — Summer Garden Food Manufacturing, formerly known as Zidian Manufa
Source: GiaRussa (Vimeo)

Name: PANGEA ORGANICS—The Story of a Truly Green Brand [score: ]
Publish Date: 2009-09-10 09:24:39
Description: Joshua Onysko, Founder + CEO of Pangea Organics. Another great talk in the AIGA Metro-North Chapter speaker series. BIO (From Pangea’s web site) It was Pangea founder and CEO Joshua Onysko’s personal devotion and commitment to inspiring social sustainability that sparked the inception of Pangea
Source: Scott Lerman (Vimeo)

Query took 0.639654159546 seconds

Results: Segment Transcript #3 (OpenCalais entities)

Name: Pallywood I – According to Palestinian Sources [score: ]
Publish Date: 2008-09-10 01:53:38
Description: The term “Pallywood” refers to the staging of scenes by Palestinian journalists in order to present the Palestinians as hapless victims of Israeli aggression. They are able to succeed in this endeavor in large part due to the credulity and eagerness of the Western press to present these images, whic
Source: Israel (Vimeo)

Name: Hamas in thier Own Voices [score: ]
Publish Date: 2009-02-02 16:23:39
Description: he video, a compilation of MEMRI TV clips that aired prior to the current Gaza crisis, includes statements by Hamas leaders calling for the annihilation of Israel and of all Jews, for death to America, and for the Islamic conquest of the world. Featured are Hamas leader Khaled Mash’al, Hamas Prime
Source: MEMRITV (Vimeo)

Name: Red State Update: Israel In Gaza [score: ]
Publish Date: 2009-02-15 21:07:59
Description: Jackie and Dunlap on the Israeli-Palestinian conflict. Get our CD “How Freedom Sounds” at Amazon, iTunes, or http://www.redstateupdate.com Distributed by Tubemogul.
Source: travis and jonathan (Vimeo)

Name: Pro-Israel Rally at Nashville Legislative Plaza – 01.11.2009 [score: ]
Publish Date: 2009-01-12 13:36:04
Description: The ICEJ joined both Christians and Jews from the local community to express our support and solidarity for Israel as it wages an unpopular war against the terrorist infrastructure of Hamas.
Source: ICEJ USA (Vimeo)

Name: Palestinian Gaza deteriorates under Hamas control [score: ]
Publish Date: 2008-04-12 02:49:17
Description: Anna Chan: Last month, our Israeli correspondents reported on the Gaza crisis. With it’s food shortages and crumbling health services under the Hamas, today’s update on Gaza talks about human rights being overidden by political interests. It’s not easy living in Gaza today. Under the control of the
Source: NTDTV (Vimeo)

Query took 1.26313090324 seconds

Results: Segment Transcript #3 (Zemanta entities)

Name: Focus Israel News 2009-09-23 [score: ]
Publish Date: 2009-10-09 18:50:55
Description: We want to complicate the picture of what is happening in Israel, because the reality is very complex. Israel is the only democracy in the region, where everybody has the right of free expression. The Watec cleantech exhibition and conference in Tel Aviv in November. The Aftonbladet Affair, that isr
Source: Fokus Israel (Vimeo)

Name: Gaza One Year Later [score: ]
Publish Date: 2009-12-07 19:36:20
Description: Filmed live at Brecht Forum in New York City on December 2, 2009: Gaza: One Year Later Norman Finkelstein on the significance of the Goldstone Report Norman Finkelstein On the morning of 27 December 2008, Israeli occupying forces launched ‘Operation Cast Lead,’ a wide-ranging military offensive ag
Source: Jonathan Shockley (Vimeo)

Name: Untitled Gaza Film [score: ]
Publish Date: 2009-11-30 16:43:29
Description: After Israel’s 27-day bombardment of Gaza in January 2009 that killed 1,400, I joined a relief delegation and stayed for two months, capturing survivors’ stories. Their voices put forth a visceral impression of the desperation and resilience that define the Gaza Strip. My use of verite style mirrors
Source: Edward Salem (Vimeo)

Query took 0.970238924026 seconds

Content APIs – Actions

Test results are using entities returned from OpenCalais or Zemanta’s entity extraction engine, and have not been curated. In the application environment, these entities will be curated by content managers, so the quality of the results below are not entirely accurate. The number of query terms and the logical operators have been adjusted to return the optimal results for each API. Obviously, there will be much more tuning once the selected API(s) are integrated into the application. The first five results are shown for each query.

Social Actions

Overview

Social Actions aggregates actions from over 60 social activism sources, including Kiva, VolunteerMatch and change.org. The SocialActions API supports filtering by “action type,” as well as creation date and specific sites. Social Actions data is also available through Zemanta , but we will be using the official Social Actions API for this test.

Service Description

Notes

  • Limited query to top 3 entities (match=any) for this test
  • Support for logical operators not supported, though the “match” parameter can be set to “any” or “all”
  • No score or relevance value returned with results
  • No date value returned with results
  • Limited documentation

Analysis
Result accuracy is low when using the match=”any” parameter; however, there are very few (if any) results when using match=”all” with the top few entities. Obtaining relevant results from the Social Actions API will take some work, though the results should improve when entities are curated prior to calling the API.

Results: Segment Transcript #1 (OpenCalais entities)

Name: Jenny Lush [Greater Good South Africa Receivers] [score: ]
Publish Date:
Description: We are a company that helps people to either get back on their feet, or to get a company up and running where a person has no money of their own to finance their idea. We are currently helping a children\\’s home in Jeffreys Bay. If you are able to…
Source: Social Actions

Name: Care Volunteers [Idealist.org Volunteer Opportunities] [score: ]
Publish Date:
Description: BackgroundEmmanuel Children’s Home resides in the town of Middelburg, halfway between Johannesburg and Cape Town in the province of Eastern Cape.Currently we have 9 children who are fostered with us due to no other suitable home being found for…
Source: Social Actions

Name: Five fellowships in Africa with community-based women’s associations [Idealist.org Volunteer Opportunities] [score: ]
Publish Date:
Description: Vital Voices, the well-known Washington-advocate for women’s rights, is seeking five AP Peace Fellows to work with five of its local partners in Cameroon, Ghana, Uganda, South Africa and Uganda. Vital Voices is building a network of local businessw…
Source: Social Actions

Name: The Last Day [WildlifeDirect] [score: ]
Publish Date:
Description: Some of my family and friends have asked why I have not contributed to my blog for months. The answer is that when the raptor expedition was over there was very little to report upon. I began this blog about the time I had frequent armed …
Source: Social Actions

Name: Honor Eudy Simelane & Corrective Rape Victims at the 2010 FIFA World Cup in South Africa. [Care2 Petitions] [score: ]
Publish Date:
Description: Please help us encourage FIFA to honor Eudy Simelane and all of the South African women who have suffered “corrective rape” at the hands of homophobic thugs who are rarely even brought to justice in South Africa. The 2010 FIFA World Cup is …
Source: Social Actions

Query took 0.529694080353 seconds

Results: Segment Transcript #1 (Zemanta entities)

Name: Jenny Lush [Greater Good South Africa Receivers] [score: ]
Publish Date:
Description: We are a company that helps people to either get back on their feet, or to get a company up and running where a person has no money of their own to finance their idea. We are currently helping a children\\’s home in Jeffreys Bay. If you are able to…
Source: Social Actions

Name: Care Volunteers [Idealist.org Volunteer Opportunities] [score: ]
Publish Date:
Description: BackgroundEmmanuel Children’s Home resides in the town of Middelburg, halfway between Johannesburg and Cape Town in the province of Eastern Cape.Currently we have 9 children who are fostered with us due to no other suitable home being found for…
Source: Social Actions

Name: Five fellowships in Africa with community-based women’s associations [Idealist.org Volunteer Opportunities] [score: ]
Publish Date:
Description: Vital Voices, the well-known Washington-advocate for women’s rights, is seeking five AP Peace Fellows to work with five of its local partners in Cameroon, Ghana, Uganda, South Africa and Uganda. Vital Voices is building a network of local businessw…
Source: Social Actions

Name: The Last Day [WildlifeDirect] [score: ]
Publish Date:
Description: Some of my family and friends have asked why I have not contributed to my blog for months. The answer is that when the raptor expedition was over there was very little to report upon. I began this blog about the time I had frequent armed …
Source: Social Actions

Name: Honor Eudy Simelane & Corrective Rape Victims at the 2010 FIFA World Cup in South Africa. [Care2 Petitions] [score: ]
Publish Date:
Description: Please help us encourage FIFA to honor Eudy Simelane and all of the South African women who have suffered “corrective rape” at the hands of homophobic thugs who are rarely even brought to justice in South Africa. The 2010 FIFA World Cup is …
Source: Social Actions

Query took 0.530394077301 seconds

Results: Segment Transcript #2 (OpenCalais entities)

Name: Repairs For Free Clothing Store [ModestNeeds PrequalifiedApplications] [score: ]
Publish Date:
Description: Our agency partners with a free clothing store that needed to have some emergency electrical work performed, because there appeared to be a rather dangerous situation involving some overhead lighting. Our funds are somewhat restricted to paying for t…
Source: Social Actions

Name: Help Free Saigon, Australia’s Last Circus Elephant [Care2 Petitions] [score: ]
Publish Date:
Description: Help End Elephant AbuseHelp Free Saigon, Australia’s Last Circus Elephant Meet Saigon. She is Australia’s last remaining circus elephant. She’s now 55 years of age and too old to perform, but she is yet to get the …
Source: Social Actions

Name: Jalal : Lebanon [Kiva] [score: ]
Publish Date:
Description: $425 of $1,200 raised. Started raising funds…
Source: Social Actions

Name: Yanis Poveda : Nicaragua [Kiva] [score: ]
Publish Date:
Description: $400 of $525 raised. Started raising …
Source: Social Actions

Name: Trades Skills Needed (Work exchange) [Idealist.org Volunteer Opportunities] [score: ]
Publish Date:
Description: Trade Skills Needed!!As an established retreat centre and spiritual community, The Salt Spring Centre of Yoga offers the ideal environment for personal growth and the practice of yoga.Do your service amid 70 acres of beautiful…
Source: Social Actions

Query took 0.443289041519 seconds

Results: Segment Transcript #2 (Zemanta entities)

Name: Request for Better Selection of Organic Foods [Care2 Petitions] [score: ]
Publish Date:
Description: We at the Burin Peninsula Environmental Reform Committee would like to help peninsula residents become more aware of food security, and the lack of Organic food availability at our local grocers. Please, for our health and safety, as…
Source: Social Actions

Name: ENVIRONMENT AND MIXED FARMING CAMPAIGN [Idealist.org Campaigns] [score: ]
Publish Date:
Description: PRESS RELEASE – THE PROMOTION OF THE POULTRY INDUSTRY IN CAMEROON On Monday 21st December, 2009, the President of SESURUDEV Association, Mr Aloysius Njie AJUA was the guest of the live programme called “Morning Safari” in the state-owned Natio…
Source: Social Actions

Name: Wildlife/Plant Life Identification/Categorization Report Volunteer [Idealist.org Volunteer Opportunities] [score: ]
Publish Date:
Description: Come join Grupo Fenix in the Solar Mountain Project. Research, Identify, Categorize, and Report on the different wildlife (birds), and plant life living in the mountain. Volunteers will be working with the women’s cooperative in Totogalp…
Source: Social Actions

Name: Coordinator for Educational Improvement in Rural Women’s Cooperative [Idealist.org Volunteer Opportunities] [score: ]
Publish Date:
Description: Come join Grupo Fenix and the Solar Women of Totogalpa. Las Mujeres Solares de Totogalpa (The Solar Women of Totogalpa, SWT) have been working together since 1999 to promote the use of renewable energy in their community. Their center is …
Source: Social Actions

Name: Reforestation Volunteers and/or Deforestation Research [Idealist.org Volunteer Opportunities] [score: ]
Publish Date:
Description: Grupo Fenix is seeking volunteers to come assist in the reforestation of the Solar Mountain. We are also seeking volunteers with educational backgrounds in environmental issues of deforestation who could research and report the current conditions of …
Source: Social Actions

Query took 0.529097080231 seconds

Results: Segment Transcript #3 (OpenCalais entities)

Name: Proposal – Help return Jewish Prayer to the Temple Mount [GiveMeaning Proposals] [score: ]
Publish Date:
Description: I have been privileged to visit our holy Temple Mount numerous times. Each time that I go to the Mount of God, I think of the verse from Psalms 24, “Who will go up to the Mount of the Lord and who will stand His holy place? One with clean hands and p…
Source: Social Actions

Name: Martin De Jesus Garcia Vallecillo : Nicaragua [Kiva] [score: ]
Publish Date:
Description: $350 of $575 raised. …
Source: Social Actions

Name: Duvon De Ftima Jimnez Rosales : Nicaragua [Kiva] [score: ]
Publish Date:
Description: $200 of $250 raised. …
Source: Social Actions

Name: California Demands Electric Car Infrastructure [Care2 Petitions] [score: ]
Publish Date:
Description: We the people want Electric Cars now, in 2010, not later.Regulations change is needed now in order to give the green light for Electric Vehicle Infrastucture here in the US and California.Denmark and Israel are aggressively pr…
Source: Social Actions

Name: 400 Animals killed by IDF in Gaza Zoo [Care2 Petitions] [score: ]
Publish Date:
Description: This zoo, located in the Zeytoun district just south of Gaza City, was almost totally and deliberately destroyed by the Israelis when they entered the Strip in December 2008. Most of the animals in that zoo had been shot at point blank range – f…
Source: Social Actions

Query took 0.415922880173 seconds

Results: Segment Transcript #3 (Zemanta entities)

Name: Proposal – Help return Jewish Prayer to the Temple Mount [GiveMeaning Proposals] [score: ]
Publish Date:
Description: I have been privileged to visit our holy Temple Mount numerous times. Each time that I go to the Mount of God, I think of the verse from Psalms 24, “Who will go up to the Mount of the Lord and who will stand His holy place? One with clean hands and p…
Source: Social Actions

Name: Martin De Jesus Garcia Vallecillo : Nicaragua [Kiva] [score: ]
Publish Date:
Description: $350 of $575 raised. …
Source: Social Actions

Name: Duvon De Ftima Jimnez Rosales : Nicaragua [Kiva] [score: ]
Publish Date:
Description: $200 of $250 raised. …
Source: Social Actions

Name: California Demands Electric Car Infrastructure [Care2 Petitions] [score: ]
Publish Date:
Description: We the people want Electric Cars now, in 2010, not later.Regulations change is needed now in order to give the green light for Electric Vehicle Infrastucture here in the US and California.Denmark and Israel are aggressively pr…
Source: Social Actions

Name: 400 Animals killed by IDF in Gaza Zoo [Care2 Petitions] [score: ]
Publish Date:
Description: This zoo, located in the Zeytoun district just south of Gaza City, was almost totally and deliberately destroyed by the Israelis when they entered the Strip in December 2008. Most of the animals in that zoo had been shot at point blank range – f…
Source: Social Actions

Query took 0.326047897339 seconds

Other APIs & Partners

Freebase

Overview

A product of Metaweb Technologies , Freebase is a structured database of Named Entities and entity types. Freebase contains over 11 million interconnected topics, all of which have strong, dereferenceable identifiers (URIs), and descriptive metadata. Topics are linked to other datasets, such as WikiPedia and the New York Times. Custom “bases” may be created for storing custom data collections. Freebase provides an open API for accessing and contributing to the database.

Service Description

DBpedia

Overview

DBpedia provides a structured dataset of Wikipedia information that can be queried and linked to other data sets. DBpedia provides RDF and JSON interfaces, and is interlinked with many other open data sets.

Service Description

Conclusions & Recommendations

The conclusions and recommendations below are based on the results of the above tests. Where possible, the Terms of Use URL for each API has been listed for legal review purposes. The Terms of Use for each API may affect the potential for use in the ViewChange.org application.

Entity Extraction APIs

In terms of quality and quantity of disambiguated Named Entities returned in these tests, Zemanta was the clear leader of the NLP API field. OpenCalais also returned highly relevant terms, but was lacking in disambiguation features. Of the other APIs tested, only AlchemyAPI returned quality, disambiguated results, though the quantity of entities returned was low.

The recommendation is to use Zemanta as the primary NLP API, supplemented with OpenCalais. While Zemanta returned the highest quality results in these tests, OpenCalais is an industry leader that will likely improve its disambiguation and Linked Data features over time. In its current incarnation, OpenCalais results will need to be manually linked with Named Entities from Freebase or DBpedia; however, this will also add some value to the ViewChange.org topic store. Using a combination of the Zemanta and OpenCalais also provides redundancy should one of the services be down at any given time.

Content APIs – Articles

Zemanta and Daylife both provide unique API features, resulting in (the potential for) higher quality results than the other article/news APIs tested here. Zemanta finds related content by analyzing text blocks (up to 8KB) instead of individual query terms. In theory, this should provide additional context for topics contained within the text. It also means that human curation of entities/topics within the application will not be taken into account by the API (though we are inquiring about this). Daylife, on the other hand, does accept entities as individual query terms, but is unique in its query weighting capabilities. Query terms can be weighted by a numeric “boost” value, focusing the results by topic relevance. Daylife also provides a higher level of descriptive metadata than the other related content APIs tested. The recommendation is to use both Zemanta and Daylife for related news/article content in the ViewChange.org application.

Content APIs – Videos

The results from the video APIs were inconsistent at best. None of the APIs tested support any advanced querying features such as textual analysis or query term weighting. Considering the vast amount of extraneous user–generated video on the web, this makes the task of finding relevant video extremely difficult. Of the APIs tested, Truveo offered the most advanced API, providing several filtering and sorting parameters. Truveo also has a large index, spanning both professional and user–contributed video sites (including YouTube), and powers video search for several large companies (some of which are partners of Link Media). The YouTube API , in comparison, indexes only its own database, though it does provide a few filters for narrowing a search to a specific category or channel. At this point, the recommendation is to explore using Truveo as the default related video API, with YouTube as a secondary choice.

Content APIs – Actions

Social Actions was the only actions API tested here, so it is clearly the recommended choice at this time. The initial test results with the Social Actions API were not ideal, though accessing the data through the Zemanta API still needs to be evaluated. Other social actions APIs are still needed for comparison.

Back to top

Tags: , ,

16 Responses to “Entity Extraction & Content API Evaluation”

Tom Tague says:

Rob:

Tom Tague from OpenCalais here. Thanks for the thoughtful write up and analysis – we’ll be reading and evaluating it carefully.

Beyond entity extraction is Viewchange looking into NLP-driven fact and event extraction? This is where we’re focusing a fair amount of our attentions recently. While we love entity extraction for it’s ability to help publishers organize and link content – we believe the real power of semantic extraction lies in understanding the “aboutness” of content – what’s happening – not just what entities are present.

Entities describe the news – facts and events are the news.

Regards,

Andraz Tori says:

Hi,

Andraz from Zemanta here.

Thank you for such a comprehensive study! While many people are using the APIs evaluated, not many take time to compare them and study them as deeply as you did!

Oh, and about related content, as mentioned in discussions, an “emphasis” parameter in Zemanta API allows for ‘manually’ guiding the discovery of content. Don’t know if you tested if that might be useful.

bye
Andraz Tori, CTO at Zemanta

Kelvin Jones says:

Fabulous write-up.

I’ve been digging around the semantic API space for a bit as I compile a list of resources, and this one will be invaluable.

I’d love to see how each API handles both non-English languages and more localized content.

Cheers,

Kelvin
——
PS: I’d love to get a copy of the test code to have a play with it.

Frederick says:

When I run the provided example texts through the Zemanta/Calais/AlchemyAPI demos, the results don’t seem to match up with the listed determinations.

Zemanta is missing lots of entity types, such as locations (towns and parks), persons, etc. This service seems more about tagging and linking to DBpedia than actual entity recognition.

When I run the listed text snippets through all these services, the results seem to indicate that EVRI, AlchemyAPI, and OpenCalais do the best at recognizing typed named entities. Also, several of these services (Calais, Alchemy) offer specialized tagging APIs, separate from named entities, and these results didn’t seem to be factored into the evaluation.

Hi Tom,

Thanks for your response. We are very interested in identifying facts and events, particularly recent events, through semantic analysis. The Calais Web Service does provide a wealth of semantic metadata through the Facts/Events and Social Tags results. What we’ve yet to figure out is how to link those results to other datasets in the Linking Open Data cloud, which is one of the goals of the ViewChange.org initiative.

Finding Linked Data URIs for recent news and events has proven difficult, leading to frequent manual creation of entries in Wikipedia and Freebase. There seems to be a considerable lag-time between news events and their inclusion in LOD datasets. We would love to learn more about how OpenCalais is bridging this gap.

Best,
Rob

Andraz:

We are experimenting with the “emphasis” parameter, as you suggested.

Thanks!

Dave Black says:

Seems to be quite a bit of variance in these Entity Extraction evaluations.

Another recent look into this space by Michael Fagan of the Bing Maps team listed OpenCalais near the bottom of the rankings, yet it’s listed as #2 here.

Here’s a link to fagan’s review:

http://faganm.com/blog/2010/01/02/1009/

Bob McCreary says:

“John Smith walked down the street in Paris, in Lamar County.”

Zemanta doesn’t get “John Smith”, or “Paris”.

Instead it comes back with “Georgia” (strange), “United States”, and “Lamar County”.

Many of the other listed APIs are detecting the entities missed by Zemanta, so I’m at a loss as to why it’s listed first. Seeing similar results for pretty much every news article I try.

Andraz Tori says:

@Frederick:
You need to use markup_limit parameter if you want to get more entities back from Zemanta. Zemanta only returns ‘known’ entities that can be tied to known web addresses, not unknown ones.

@Bob McCreary:
Zemanta isn’t designed to run on such short sentences (albeit it works most of the time). Try it with larger article. Also you seem to be looking at “keywords” instead of “markup” part of the response.

bye
Andraz Tori, CTO at Zemanta

Well done Andraz!
Congratulations to you and your team.

Mitul says:

Very nice article comparing different entity extractor and content apis.

Just to mention, Kosmix.com builds topic pages for any entity and might be worth mentioning as a content provider. For example, http://www.kosmix.com/topic/akon or http://www.kosmix.com/topic/Splice_(film)

Also, Kosmix provides an api to recognize, disambiguate, and link entities in a document. For example, http://www.healthboards.com/boards/showthread.php?t=748185

Just to mention, I work at Kosmix, and appreciate your article comparing different products.

I’ve posted the test tool source code on GitHub:

http://github.com/robdiciuccio/Simple-API-Test-Tool

Hello and thanks for including AlchemyAPI in your evaluation!

Curious which NLP APIs from Alchemy were used in this comparison: AlchemyAPI provides both a Zementa-style “Concept Tagging” API and a traditional “Entity Extraction” API.

If you’re receiving a low number of entity results, you’re likely using only one of our APIs. For the maximum number of semantically linked results (for Keywords/Tags AND named Entities) be sure to use both Named Entity Extraction ( http://www.alchemyapi.com/api/entity/ ) and Concept Tagging ( http://www.alchemyapi.com/api/concept/ ).

AlchemyAPI’s Concept Tagging API returns disambiguated and linked-data-enhanced results for a wide variety of concepts and named entity types. When comparing against solutions such as Zementa you’ll find this API provides a more “apples to apples” comparison.

Thanks for this very insightful analysis. I repeated your tests on many different data sets, and can fully confirm your results. My focus was mainly on the Linked Data aspect, Zemanta and AlchemyAPI both do a really good job there.

@tomayac

This its really a greate article about semantic services. We are also comparing entity extraction services right know – maybe you like our little Flash/Flex tool where you see all results at once:
http://www.veeeb.de/blog/news/entity-extraction-alchemyapi-evri-opencalais/
Have fun!
Christoph Diefenthal

John Lehmann says:

While everyone’s throwing their hat in the ring, I thought Extractiv might as well. :)

Extractiv’s novelty is that we run our NLP on a web crawler, so you can transform the open unstructured web into structured semantic output. This is great when you want to discover information from the web as opposed to process collections of documents you already have. For the latter case, we provide a REST-based on-demand service.

We currently provide over 150 entity types, relations, and more, in output formats including JSON and RDF. I’m with Tom Tague when he said “Entities describe the news – facts and events are the news.” We like entities, but want to provide a lot more than that. We’re new, so bear with us as we still have some additional services to roll out, including entity linking, which should be within a few weeks.

To usher us in as a late contender to this post, I’ve put URLs by which you can get live results on these 3 transcripts (or any URL), which hits our demo server. I promise I didn’t doctor the results. :)

http://rest.extractiv.com/extractiv/?url=http://rest.extractiv.com/sample_docs/viewchange1.txt&output_format=html_viewer#
http://rest.extractiv.com/extractiv/?url=http://rest.extractiv.com/sample_docs/viewchange2.txt&output_format=html_viewer#
http://rest.extractiv.com/extractiv/?url=http://rest.extractiv.com/sample_docs/viewchange3.txt&output_format=html_viewer#

website: http://www.extractiv.com
blog: http://blog.extractiv.com
docs: http://wiki.extractiv.com
twitter: http://twitter.com/#!/Extractiv (@extractiv)