The 17th International World Wide Web Conference (April 21 - 25, 2008, Beijing, China) has a workshop session titled “Natural Language Processing Challenges in the Information Explosion Era” (NLPIX 2008). This workshop is calling for papers dealing with Natural Language Processing (NLP) issues on multiple topics, including trust/credibility evaluation of Web documents using NLP. The deadline for paper submissions is January 26, 2008.

Johnson, Kaye, Bichard, and Wong (2007) surveyed 1,399 internet users online during a four-week period surrounding the 2004 presidential election day (October 19 to November 16). The study applied uses and gratification theory to study credibility, which purportedly had not been widely used by credibility researchers. Amongst the hypotheses tested were whether motivations for using blogs and whether reliance on blogs for political information significantly predicted perception of blog credibility. Motivations were measured in terms of: 1) information seeking/media checking (searching about current issues), 2) convenience (using blogs because it’s more convenient and easier to reach than traditional media or other online sources), 3) personal fulfillment (fulfilling emotional needs and gathering information to use in discussion with others), 4) political surveillance, and 5) social surveillance (learning about others’ opinions).

Credibility was measured in terms of believability, fairness, accuracy, in-depth, and a combined index. After controlling for political and demographic variables, reliance on blogs was found to be a stronger predictor than motivation; and political information seeking was the strongest motivational predictor. In addition, participants found blogs to be higher in credibility for depth of information than traditional media or other online sources (p <.001, two tailed), albeit lowest in credibility for fairness of information. You can read the full article from this 2007 Journal of Computer Mediated Communication here.

Nick Haley’s YouTube “commercial” of iPod Touch was spotted by Apple’s ad executives. They bought his ad, which now airs on TV. While this piece of “news” is a few weeks old, it is worth mentioning here because it is an example of how social media can support anyone who is not an authority figure to be known based on skills and creativity per se. This example might underscore the relevance of the multiple dimensions of credibility pertaining to content, and perhaps raises the question of whether and how newness and creativity of content affect perceived credibility in new forms of social media–or whether these dimensions are about of “credibility” or not.

Nick Haley’s YouTube version of the iPod Touch commercial can be seen here. Additional news stories about it can be found in the New York Times and Wired blog.

A study of information evaluation and coping styles of people living with HIV/AIDS reports that “assigning higher credibility to unfounded Internet information was predicted by lower incomes, less education, and avoidant copying styles.” You can read Kalichman, Cherry, Cain, Weinhardt, et al.. (2006)’s study abstract in this PubMed entry.

On August 31, 2007, Google News announced a feature that will remove duplicate results from its aggregated links to articles published on news outlets’ websites. This is the outcome of a licensing agreement with the Agence France Press (AFP), the Associated Press (AP), the U.K. Press Association, and the Canadian Press. On the one hand, such agreements relieve Google from lawsuits against copyright infringement on the wire services; on the other hand, Google is able to provide more original content to users and more direct access and awareness of source origin. Becoming removed from or having difficulty tracing the origin of a source is a concern that current social media models of production raise. I wonder how Google News’ model may lead to other creative approaches to easily tracing source origin in social media outlets.

For more information about the relationship between news wires services, online newspaper outlets, and Google News, see this InfoWorld article.

This past August 20, I wrote blog post introducing the Wikipedia Trust Coloring Tool from the UC Santa Cruz’s Wiki Lab. Since then, the tool has been featured on slashdot and Wired. To this blogpost on slashdot, people have responded with their thoughts about the tool. Someone suggested to create a wikipedia entry about the algorithm and run the tool on the entry to assess its trustworthiness. Another person discussed the potential that those who reveal misdeeds are judged less trustworthy because their entries are likely to be edited out. In this Wired article, Bergstein acknowledges that the tool is not foolproof for the reasons just mentioned, but reveals “interesting insights”. Bergstein reports that the tool’s developer, Luca de Alfaro, hopes to work with the Wikimedia Foundation to make the tool a real-time option for users. However, no official decisions have been made. How people react to such tool and appropriate it might determine the tool’s future trajectory.

An excerpt from an article by Dr. Rieh, Soo Young conceptualizes credibility and related terms. Rieh is Assitant Professor of the School of Information at the University of Michigan. The excerpt can be found here.

The CredibilityCommons’ 2006-2007 Annual Report has been added to the publications page on the CredibilityCommons website. The report summarizes the CredibilityCommons’ progress in the development of tools and technology, research, collaboration, and dissemination.

According to an article in globeandmail.com by Anita Chang, Associated Press, August 22, 2007, “Blog service providers in China are “encouraged” to register users with their real names and contact information, according to a new government document that tones down an earlier proposal banning anonymous online blogging”. Some blog service providers signed the “self discipline pledge” of the Internet Society of China, which according to InfoWorld, include Yahoo and Microsoft. According to Chang, the government document detailing this issue stipulated guidelines and not requirements. It would be interesting to see what the ISOC’s reaction is if guidelines are not followed, something the article does not mention. If only policing online identity could solve some credibility issues! You can read the rest of the article here. Read also how Reporters Without Borders critized the Chinese government in this InfoWorld article.

According to a Reuters report in InformationWeek (August 21, 2007), Ask.com will capitalize on the aggregate behavior of different type of users to improve search results. While major search systems all analyze the collective online surfing behavior of users, Ask’s “affinity-group” approach to collective search will apparently include more specific user characteristics. In the context of credibility research, what difference does Ask’s type of collective search results have for people’s assessment of information credibility? Read more about this report here.

Over 50-year-old ’silver surfers’ are increasing in numbers on internet use, including using social media and social networks, such as Eons–and are creating a market opportunity not yet widely tapped. According to a Communications Market Report from the UK’s Ofcom (Office of Communications, an independent regulator and competition authority for the UK communication industires), in the UK, “nearly 30% of total time spent on the interenet is accounted for by over-50s”. While I have found no information about usage statistics on social news media, it is likely that assessing the credibility on social news media sites is an issue spanning different generations. [Content in this post first appeared in this Wired Blog Network post.]

The Untiversity of California, Santa Cruz’ Wiki Lab show their Wikipedia trust coloring demo. The text in a Wikipedia page is colored according to its contributor’s reputation (See what each color means). The reputation of Wikipedia authors is computed according to how long an author’s contribution lasts in the Wikipedia. “Specifically, authors whose contributions are preserved, or built-upon, gain reputation; authors whose contributions are undone lose reputation.” Details of the algorithm are presented in their paper A Content-Driven Repuration System for the Wikipedia. [The UCSC Wiki Lab was first mentioned in this article from The Jerusalem Post.]

Since Wired News first wrote about WikiScanner last week, Internet users have spotted plenty of interesting changes to Wikipedia by people at nonprofit groups and government entities like the Central Intelligence Agency. Many of the most obviously self-interested edits have come from corporate networks. [Copied from this article in The New York Times.]

The brainchild of Cal Tech computation and neural-systems graduate student Virgil Griffith — offers users a searchable database [called WikiScanner] that ties millions of anonymous Wikipedia edits to organizations where those edits apparently originated, by cross-referencing the edits with data on who owns the associated block of internet IP addresses. [Copied from article in Wired, Aug. 14, 2007.] Wired suggests you try out the software and submit your vote on wikipedia submissions.

The AOL Developer Network provided a status update (August 14, 2007) about the third party OpenIDs AOL supports. 10 third party providers are supported, and AOL is accepting requests to include more. Though OpenIDs are not a new thing, its recent mentioning in the claimID blog and AOL Developer Network has generated new discussions worth mentioning in the context of ‘credibility’. What new developments can OpenID lead to in terms of facilitating people’s assessment of information? To learn more about OpenID check out this article by Fred Stuztman.

An article in the InformationWeek, August 15, 2007, presents perspectives and reports from educators and researchers about the role of Hollywood in scientific and technological literacy. Mentioned in the article are a paper by UCF professor Costas J. Efthimiou and former UCF physics chair R. A. Llewellyn titled “Hollywood Blockbusters: Unlimited Fun But Limited Science Literacy”, and a 2006 report by the Association of Computing Machinery on Globalization and Offshoring of Software. You can read the InformationWeek article here.

Taiwanese academics debate the credibility of information on Wikipedia. Some critique its utility for education while others appraise the value of a “wiki” model. Associate professor S. T. Huang of the National Pingtung University of Science and Technology says they have been using the wiki model to accumulate and preserve “disappearing local knowledge”. Researchers at Academic Sinica have also been inspired by the wiki model, and are in the process of compiling the Taiwanese edition of the Encyclopedia of Life (EOL, 生命大百科), in which they aim “to include information on at least 80 percent of species that can only be found in Taiwan.” You can read more about this in the Taipei Times.

On August 7, 2007, Google announced an experimental feature on Google News which will display comments from people and organizations who are actual participants in the news stories discussed.  Google’s aim is to “help enhance the news experience for readers” by providing both professionally written articles and personal views. However, critics question the practicalities and implications of such a project on problems such as verifying a source’s identity or screening for inaccurate statements. You can read more on this New York Times aritlce and on Google News blog.

The John D. and Catherine T. MacArthur Foundation announced on August 14, 2007 a public competition that will award $2 million in funding to emerging leaders, communicators, and innovators shaping the field of digital media and learning.  The competition is administered by HASTAC (the Humanities, Arts, Science and Technology Advanced Collaboratory) and is part of MacArthur’s $50 million Digital Media and Learning initiative that aims to help determine how digital technologies are changing the way young people learn, play, socialize, and participate in civic life. Awards will be given in two categories: Innovation Awards (from $100,000 to $250,000) and Knowledge Networking Awards (from $30,000 to $75,000). You can access the competition site, or read the full press release on the MacArthur site and on the Chronicle of Higher Education.

Researchers at Microsoft Research and the University of California, Davis have performed a comprehensive analysis of search engine spam techniques that may be useful in identifying and removing these sites from search engine results. You can also read the original paper, presented at the WWW 2007 conference.

Credibility Commons project manager Shaun Kane gave two presentations this month about the Credibility Commons and other issues related to Internet credibility.

On January 12th, Shaun gave a presentation about the Commons projects at Microsoft Research. On January 26th, Shaun and PI Mike Eisenberg presented their work as part of the University of Washington’s Information School Research Conversations.

Slides from both presentations are linked to this post. We welcome your comments.

MSR presentation
iSchool Research Conversation

We have posted two new pieces of software that you can try out. CredibilityTalk is a social web site where users can rate and discuss the credibility of web sites. The Credibility Toolbar is a plugin for the Firefox browser that allows users to access and add ratings to the CredibilityTalk database.
You can access these tools on our Software page, or by visiting http://credibilitycommons.org/toolbar/. We hope that you will try out these tools and share your comments with us.