Contextual Link Analysis

Abstract:

A web page (or author) is not equally authoritative on all topics. However, traditional link analysis techniques consider only a single measure of authority, with little or no concern for the topic or community from which that authority is derived. This research finds ways to determine the context of recognized authority, and to incorporate that context as topicality or community recognition into more accurate automated analyses of reputation. This project further improves link analysis by distinguishing between popularity, quality, and authority. By integrating such analysis into fundamental link analysis techniques, search result quality improvements are demonstrated using human relevance and quality judgements. This contextual link analysis provides new opportunities to significantly improve measurements of authority and thus search engine result rankings (as well as topical rankings for other purposes), and to better understand the topical structure of interlinked document networks such as the Web and scholarly publications. The project additionally incorporates educational and outreach activities, including a program to stimulate undergraduate interest and experience in computer science research in area schools, while promoting interaction among nearby faculty.

• Faculty: Brian D. Davison
• Students: Lan Nie, Xiaoguang Qi, Jian Wang
• Former members: Baoning Wu (PhD, 2007), Shruti Bhandari (MS, 2008), YaoShuang Wang (MS, 2008)

Primary publications:

• X. Qi and B. D. Davison. (2008) Classifiers without Borders: Incorporating Fielded Text from Neighboring Web Pages. To be published in Proceedings of the 31st Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, Singapore, July.

• L. Nie and B. D. Davison. (2008) Separate and Inequal: Preserving Heterogeneity in Topical Authority Flows. To be published in Proceedings of the 31st Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, Singapore, July.

• L. Nie, B. D. Davison and B. Wu. (2007) From Whence Does Your Authority Come? Utilizing Community Relevance in Ranking. In Proceedings of the Twenty-Second Conference on Artificial Intelligence (AAAI-07), pages 1421-1426, Vancouver, Canada, July 22-26.

• L. Nie, B. Wu and B. D. Davison. (2007) Winnowing Wheat from the Chaff: Propagating Trust to Sift Spam from the Web. In Proceedings of the 30th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, pages 869-870, Amsterdam, July.

• L. Nie, B. D. Davison and B. Wu. (2007) Ranking by Community Relevance. In Proceedings of the 30th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, pages 873-874, Amsterdam, July.

• L. Nie, B. Wu and B. D. Davison. (2007) A Cautious Surfer for PageRank. In Proceedings of the 16th International World Wide Web Conference (WWW), pages 1119-1120, Banff, Canada, May 9-12.

• L. Nie, B. D. Davison and X. Qi. (2006) Topical Link Analysis for Web Search. In Proceedings of the 29th Annual International ACM SIGIR Conference on Research & Development in Information Retrieval, pages 91-98, Seattle, WA, August.

This material is based upon work supported by the National Science Foundation under Grant No. 0545875. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.