Modeling and Inferring Searcher Intent by Mining User Interactions

Inferring searcher intent is a central problem in information retrieval and web search: for effective ranking and result presentation, the search engine must know what the user is looking for. Yet, expressing a searcher information need currently relies on entering the .right. search keywords, which can require multiple rounds of trial-and-error from the searcher. The goal of this project is to develop effective methods for a search engine to automatically infer searcher intent and information needs from the searcher interactions and behavior data. Specifically, the project addresses two main challenges of search intent inference: developing accurate and robust models of searcher intent and behavior, and exploiting these models to infer search intent for each individual user. This project significantly advances previous efforts on implicit feedback and search modeling, by considering a wide range of user interaction and contextual features, and by developing novel techniques for mining and exploiting these signals to improve web search and information access.

News

7/15/2011: Paper on remote studies of search result examination to appear in SIGIR 2011. More details here: Paper, citation information, demo, and data.

7/1/2011: Paper on modeling types of search success via search contest data to appear in SIGIR 2011. More details here: Paper, citation information, data, and code.

7/1/2010: New Scientist wrote a brief news blurb and a feature story about our SIGIR 2010 paper.

Tools and Datasets

Under construction

Faculty

Eugene Agichtein, Mathematics and Computer Science, Emory University

Students

Current and past collaborators

Evgeniy Gabrilovich (Yahoo! Research)
Joan A. Smith (Emory University Libraries)
Phillip Wolff, Department of Psychology, Emory University
Charles L Clarke, University of Waterloo

Publications

Ready to Buy or Just Browsing? Detecting Web Searcher Goals from Interaction Data,
Q. Guo and E. Agichtein, Proc. of SIGIR 2010
Towards Predicting Web Searcher Gaze Position from Mouse Movements,
Q. Guo and E. Agichtein, In Extended Abstracts, Proc. of CHI 2010
Query Ambiguity Revisited: Clickthrough Measures for Distinguishing Informational and Ambiguous Queries,
Y. Wang and E. Agichtein, in Proc. of (NAACL-HLT), 2010
Exploring Searcher Interactions for Distinguishing Types of Commercial Intent (poster),
Q. Guo, and E. Agichtein, in Proc. of WWW 2010
In the Mood to Click? Towards Inferring Searcher Receptiveness to Advertising
Q. Guo, E. Agichtein, C. Clarke and A. Ashkan
Proc. of the ACM/IEEE International Conference on Web Intelligence (WI), 2009
EMU: The Emory User Behavior Modeling System for Automatic Library Search Evaluation: Preliminary Results (poster),
Qi Guo, Ryan P. Kelly, Selden Deemer, Arthur Murphy, Joan A. Smith, and Eugene Agichtein
Proc. of the 9th Joint Conference on Digital Libraries (JCDL), 2009
Classifying and Characterizing Query Intent in Sponsored Search,
Azin Ashkan, Charles Clarke, Eugene Agichtein, and Qi Guo,
Proc. of the 31st European Conference on Informational Retrieval (ECIR), 2009
Understanding "Abandoned" Ads: Towards Personalized Commercial Intent Inference via Mouse Movement Analysis,
Qi Guo, Eugene Agichtein, Charles Clarke and Azin Ashkan,
SIGIR 2008 Workshop on Information Retrieval in Advertising (IRA), 2008
Characterizing Query Intent From Ad Clickthrough Data,
Azin Ashkan, Charles Clarke, Eugene Agichtein and Qi Guo,
SIGIR 2008 Workshop on Information Retrieval in Advertising (IRA), 2008
Exploring Client-Side Instrumentation for Personalized Search Intent Inference: Preliminary Experiments,
Qi Guo and Eugene Agichtein,
AAAI 2008 Workshop on Intelligent Techniques for Web Personalization and Recommender Systems (ITWP 2008)
Exploring Mouse Movements for Inferring Query Intent (poster)
Qi Guo and Eugene Agichtein
Proc. of the 31st Annual International ACM SIGIR Conference (SIGIR 2008)

Research supported by: