Modeling and Inferring Searcher Intent by Mining User Interactions

Website for National Science Foundation Award Number: IIS 1018321

Last Updated: October 2017

Eugene Agichtein

Associate Professor
Department Mathematics and Computer Science
Emory University
Office:  400 Dowman Drive, Suite W401, Atlanta, Georgia    30322,  USA
Telephone: (404) 727-7962
Fax: (404) 727-5611
Email: eugene@mathcs.emory.edu

Project summary: 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 and Updates

Award Information

This website is based upon work supported by the National Science Foundation under Grant Number 1018321. 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.

  • Award Number: IIS-1054199
  • Duration: September 1, 2010 to August 31, 2013 (with no-cost extension until August 2014)
  • Award Amount: $500,000.00
  • Award title: III: Small: Modeling and Inferring Searcher Intent by Mining User Interactions

Project Background

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.

Project Goals

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.

Project Impact

The techniques developed in this project are expected to make web search and information access more intuitive and effective for millions of users through collaboration with major search engine companies. Additional broader impacts will be achieved through domain-specific applications of the developed techniques, ranging from improved library search to web-based diagnostics of cognitive impairment.

Broad Impact and Outreach Activities

Part of research results in this project have been used in information retrieval courses (CS 571) and undergraduate CS 190, CS325, and CS370 courses.

Undergraduate students were also involved in the project as independent study participants and research assistants.

Educational Materials:

Tools and Datasets

  • Code and Data for Improving Search Result Summaries by Using Searcher Behavior Data [Ageev, Lagun, and Agichtein, SIGIR 2013].
  • Code and Data for the UFindIt game for modeling search success [Ageev, Guo, Lagun, and Agichtein, SIGIR 2011].
  • Code and data for ViewSer, a system for remote studies of search result examination [Lagun and Agichtein, SIGIR 2011]

Participants

  • Eugene Agichtein Principal Investigator
  • PhD Students:
    Dr. Dmitry Lagun,  graduated Summer 2014, first position: Research Scientist at Google.
    Dr. Qi Guo, graduated  in Fall 2012, First position: Applied Scientist at Bing.com
    Dr. Denis Savenkov, graduated Spring 2017, first position: Applied Scientist at Facebook
    Dr. Qiaoling Liu, graduated Fall 2014, first position: Data Scientist, CareerBuilder.com
  • Visitors:
    Dr. Mikhail Ageev,  Visiting Scholar, Moscow State University (Winter 2011, Winter 2013)
  • Undergraduate research assistants/programmers:
    Spencer Caroll, Undergradute Research Assistant
    William Savoie, Undergradute Research Assistant
    Suzie Noh, Undergaduate Research Assistant
    Zelma Gist, Undergaduate Research Assistant
    Ilya Shats, Undergaduate Research Assistant
    Josh Weinstock, Undergaduate Research Assistant
    Haojian Jin, Graduate Research Assistant/programmer (MS)

Publications

2013-2015

  • Lagun, D. and Agichtein, E., 2015, August. Inferring searcher attention by jointly modeling user interactions and content salience. In Proceedings of the 38th International ACM SIGIR Conference on Research and Development in Information Retrieval (pp. 483-492).
  • Lagun, D., Ageev, M., Guo, Q. and Agichtein, E., 2014, February. Discovering common motifs in cursor movement data for improving web search. In Proceedings of the 7th ACM international conference on Web search and data mining (pp. 183-192). best student paper award
  • Liu, Q., Jurczyk, T., Choi, J. and Agichtein, E., 2015, March. Real-time community question answering: Exploring content recommendation and user notification strategies. In Proceedings of the 20th International Conference on Intelligent User Interfaces (pp. 50-61).
  • Lagun, D. and Agichtein, E., 2014, July. Effects of task and domain on searcher attention. In Proceedings of the 37th international ACM SIGIR conference on Research & development in information retrieval (pp. 1087-1090). ACM.
  • D. Lagun, C. Manzanares, E. Buffalo, S. Zola, and E. Agichtein. Predicting Cognitive Decline from Eye Movement Shapelets, ICML 2013 Workshop on “Machine Learning for Health”, July 2013, Atlanta, GA
  • Savenkov, D. and Agichtein, E., 2014, July. To hint or not: exploring the effectiveness of search hints for complex informational tasks. In Proceedings of the 37th international ACM SIGIR conference on Research & development in information retrieval (pp. 1115-1118).

2012-2013

  • Towards task-based snippet evaluation: preliminary results and challenges M Ageev, D Lagun, E Agichtein, SIGIR 2013 workshop on Modeling User Behavior (MUBE 2013)
  • Mining touch interaction data on mobile devices to predict web search result relevance, Q Guo, H Jin, D Lagun, S Yuan, E Agichtein, in Proc. of SIGIR 2013
  • Improving search result summaries by using searcher behavior data, M Ageev, D Lagun, E Agichtein, in Proc. of SIGIR 2013.
  • Workshop on health search and discovery: helping users and advancing medicine, RW White, E Yom-Tov, E Horvitz, E Agichtein, W Hersh, Workshop at SIGIR 2013.
  • Characterizing eye movement as a predictor of cognitive decline: Applying machine learning to improve prediction accuracy, D Lagun, C Manzanares, E Buffalo, E Agichtein, S Zola, Alzheimer’s & Dementia: The Journal of the Alzheimer’s Association (abstract), 2013
  • Towards estimating web search result relevance from touch interactions on mobile devices, Q Guo, H Jin, D Lagun, S Yuan, E Agichtein, CHI’13 Extended Abstracts on Human Factors in Computing Systems, 2013
  • The Answer is at your Fingertips: Improving Passage Retrieval for Web Question Answering with Search Behavior Data. M Ageev, D Lagun, E Agichtein, in EMNLP 2013
  • Predicting web search success with fine-grained interaction data, Q Guo, D Lagun, E Agichtein, CIKM 2013
  • Qi Guo, Dmitry Lagun, Eugene AgichteinPredicting web search success with fine-grained interaction data. CIKM 2012. PDF

2011-2012

Research supported by: National Science Foundation Award Number: IIS 1018321 and gifts from Microsoft Research and Yahoo! Research.