1. Leskovec, Dumais, and Horvitz, "Web Projections: Learning from Contextual Subgraphs of the Web." In WWW 2007 (10 pages, pdf) Read this one first; skim the others if necessary.
2. Vassilvitskii and Brill, "Using web-graph distance for relevance feedback in search." In SIGIR 2006 (8 pages, acm link, pdf)
3. Diaz, "Performance prediction using spatial autocorrelation." In SIGIR 2007. (8 pages, acm access, CIIR access) Discusses another way to induce a graph on the top ranked documents, and how a property of that graph can be used to good effect.

1. ...

IR seminar, February 19, 2008

Web projections, Sue Dumais (Microsoft) leading

February 19th is a Tuesday (but a Monday class schedule).

Note room and time change: We will meet from 9:00-11:00 in CS151. Breakfast sort of stuff available starting around 8:30am

Today's discussion focuses on a WWW 2007 paper on "Web Projections". This paper presents some things that can be predicted about Web queries, based on properties of the Web subgraph induced by the top ranked documents.

The discussion will be led by Sue Dumais, Principal Researcher at Microsoft Research.

Required readings

1. Leskovec, Dumais, and Horvitz, "Web Projections: Learning from Contextual Subgraphs of the Web." In WWW 2007 (10 pages, pdf)

Recommended readings

1. Diaz, "Performance prediction using spatial autocorrelation." In SIGIR 2007. access Discusses another way to induce a graph on the top ranked documents, and how a property of that graph can be used to good effect.

Other related readings (feel free to add)

1. Kleinberg, "Authoritative sources in a hyperlinked environment." Journal of the ACM, 46(5):604-632, 1999. This paper presents the HITS algorithm. (pdf preprint, 34 pages)
2. Page et al., "The PageRank^? citation ranking: Bringing order to the Web." Technical report, Stanford, 1998. Discusses pagerank. (access, 17 pages)
3. ...