Computing trust from revision history

Publication (help)
Computing trust from revision history
Authors:	Honglei Zeng, Maher A. Alhossaini, Li Ding, Richard Fikes, Deborah L. McGuinness [edit item]
Citation:	Proceedings of the 2006 International Conference on Privacy, Security and Trust: Bridge the Gap Between PST Technologies and Business Services  : . 2006.
Publication type:	Conference paper
Peer-reviewed:	Yes
Database(s):
DOI:	10.1145/1501434.1501445.
Google Scholar cites:	Citations
Link(s):	Paper link
Added by Wikilit team:	Added on initial load
Search
Article:	Google Scholar BASE PubMed
Other scholarly wikis:	AcaWiki Brede Wiki WikiPapers
Web search:	Bing Google Yahoo! — Google PDF
Other:
Services
Format:	BibTeX

Computing trust from revision history is a publication by Honglei Zeng, Maher A. Alhossaini, Li Ding, Richard Fikes, Deborah L. McGuinness.

Abstract

A new model of distributed, collaborative information evolution is emerging. As exemplified in Wikipedia, online collaborative information repositories are being generated, updated, and maintained by a large and diverse community of users. Issues concerning trust arise when content is generated and updated by diverse populations. Since these information repositories are constantly under revision, trust determination is not simply a static process. In this paper, we explore ways of utilizing the revision history of an article to assess the trustworthiness of the article. We then present an experiment where we used this revision history-based trust model to assess the trustworthiness of a chain of successive versions of articles in Wikipedia and evaluated the assessments produced by the model.

Research questions

"In the work reported in this paper, we have explored the hypothesis that revision information can be used to compute a 1www.wikipedia.com measure of trustworthiness of revised documents. Based on that hypothesis, we developed a revision history-based trust model for computing and tracking the trustworthiness of the documents in collaborative information repositories. We represent our trust model in a dynamic Bayesian network (DBN) because a DBN is a powerful framework for modeling processes that evolve dynamically over time, which in our case, are ever changing articles."

Research details

Topics:	Featured articles, Vandalism, Reader perceptions of credibility, Computational estimation of trustworthiness [edit item]
Domains:	Computer science [edit item]
Theory type:	Design and action [edit item]
Wikipedia coverage:	Main topic [edit item]
Theories:	"Undetermined" [edit item]
Research design:	Experiment [edit item]
Data source:	Experiment responses, Wikipedia pages [edit item]
Collected data time dimension:	Cross-sectional [edit item]
Unit of analysis:	Article [edit item]
Wikipedia data extraction:	Dump [edit item]
Wikipedia page type:	Article [edit item]
Wikipedia language:	English [edit item]

Conclusion

"We introduced the concept of revision historybased trust and developed a dynamic Bayesian network trust model that utilized rich revision information in Wikipedia. Our experiments showed promising results, even though we made several simplifying assumptions in this work. We showed an evaluation method inWikipedia based on its feature articles, clean-up articles and the levels of author editing privileges. Our work provided a methodology for comparing and evaluating future computational trust algorithms. Based on our DBN model, we believe the reasons for Wikipedia being generally trustworthy are: (1) most Wikipedia authors seem to have good intentions (there are only 1:3% blocked authors); (2) Wikipedia administrators have the responsibility and authority to settle disputes, prevent vandalism, and block inappropriate authors. While there are a small number of administrators (0:09%), they have made much larger contributions to Wikipedia, for example, 29:4% revisions of featured articles were made by administrators in our experiments, according to Table 1; (3) Wikipedia maintains a complete revision history of articles from which a previous content modification can be easily reverted. The benefits of revision trust to Wikipedia users are significant. Visualization of the computed trust values may help users to decide what information they should trust. Users may also have the option to view the most trustworthy version of an article, in addition to the most recent one. Furthermore, revision trust can improve Wikipedia’s quality control process; for example, our model provides an appealing approach to monitoring changes in trustworthiness and thereby providing timely notifications of vandalism and other forms of malicious content modifications."

Comments

"WIkipedia is generally trustworthy and visualization of the computed trust values may give users the option to view the most trustworthy version of an article, which can improve Wikipedia’s quality control process; by providing timely notifications of vandalism and other forms of malicious content modifications."