Comparing relevance feedback algorithms for web search

We evaluate three different relevance feedback (RF) algorithms, Rocchio, Robertson/Sparck-Jones (RSJ) and Bayesian, in the context of Web search. We use a target-testing experimental procedure whereby a user must locate a specific document. For user relevance feedback, we consider all possible user choices of indicating zero or more relevant documents from a set of 10 displayed documents. Examination of the effects of each user choice permits us to compute an upper-bound on the performance of each RF algorithm. We find that there is a significant variation in the upper-bound performance of the three RF algorithms and that the Bayesian algorithm approaches the best possible. Categories and Subject Descriptors H.3.3 Information Search and Retrieval ? relevance feedback General Terms Algorithms, Performance, Experimentation Keywords Relevance Feedback, Web Search, Evaluation

• Type: Conference
• Year: 2005
• Where: WWW
• Authors: Vishwa Vinay, Kenneth R. Wood, Natasa Milic-Frayling, Ingemar J. Cox
• Keywords: Internet Technology, Keywords Relevance Feedback, Possible User Choices, User Relevance Feedback, WWW 2005