Dorota Glowacka
Giulio Jacucci
ABSTRACT
Techniques for both exploratory and known item search tend to direct only to more specific subtopics or individual documents, as opposed to allowing directing the exploration of the information space. We present an interactive information retrieval system that combines Reinforcement Learning techniques along with a novel user interface design to allow active engagement of users in directing the search. Users can directly manipulate document features (keywords) to indicate their interests and Reinforcement Learning is used to model the user by allowing the system to trade off between exploration and exploitation. This gives users the opportunity to more effectively direct their search nearer, further and following a direction. A task-based user study conducted with 20 participants comparing our system to a traditional query-based baseline indicates that our system significantly improves the effectiveness of information retrieval by providing access to more relevant and novel information without having to spend more time acquiring the information.
Supplemental Material
- Alonso, O., Baeza-Yates, R., and Gertz, M. Exploratory search using timelines. In SIGCHI 2007 Workshop on Exploratory Search and HCI Workshop, no. 1 (2007).Google Scholar
- Auer, P. Using confidence bounds for exploitation-exploration trade-offs. Journal of Machine Learning Research 3 (2002), 397--422. Google ScholarDigital Library
- Auer, P., Hussain, Z., Kaski, S., A. Klami, J. Kujala, Laaksonen, J., Leung, A., Pasupa, K., and Shawe-Taylor, J. Pinview: Implicit feedback in content-based image retrieval. In JMLR Workshop and Conference Proceedings: Workshop on Application of Pattern Analysis, vol. 11 (2010), 51--57.Google Scholar
- Balabanović, M. Exploring versus exploiting when learning user models for text recommendation. User Modeling and User-Adapted Interaction 8 (1998), 71--102. Google ScholarDigital Library
- Baldonado, M., and Winograd, T. Sensemaker: an information-exploration interface supporting the contextual evolution of a user's interests. In Proc. of SIGCHI (1997), 11--18. Google ScholarDigital Library
- Brooke, J. Sus - a quick and dirty usability scale. Usability Evaluation in Industry 189 (1996), 194.Google Scholar
- Chapelle, O., and Li, L. An empirical evaluation of thompson sampling. In Proc. of NIPS, vol. 25 (2011).Google Scholar
- Chau, D., Kittur, A., Hong, J., and Faloutsos, C. Apolo: making sense of large network data by combining rich user interaction and machine learning. In Proc. of SIGCHI (2011), 167--176. Google ScholarDigital Library
- Głowacka, D., and Shawe-Taylor, J. Content-based image retrieval with multinomial relevance feedback. In JMLR Workshop and Conference Proceedings: 2nd Asian Conference on Machine Learning, vol. 13 (2010), 111--125.Google Scholar
- Havre, S., Hetzler, E., Perrine, K., Jurrus, E., and Miller, N. Interactive visualization of multiple query results. In Proceedings of the IEEE Symposium on Information Visualization (2001), 105. Google ScholarDigital Library
- Hutchins, E. L., Hollan, J. D., and Norman, D. A. Direct manipulation interfaces. Hum.-Comput. Interact. 1, 4 (Dec. 1985), 311--338. Google ScholarDigital Library
- Joachims, T., Freitag, D., and Mitchell, T. Webwatcher: A tour guide for the world wide web. In Proc. of IJCAI (1997).Google Scholar
- Käki, M. Findex: search result categories help users when document ranking fails. In Proc. of SIGCHI (2005), 131--140. Google ScholarDigital Library
- Kelly, D., and Fu, X. Elicitation of term relevance feedback: an investigation of term source and context. In Proc. of SIGIR (2006), 453--460. Google ScholarDigital Library
- Kules, B., Wilson, M., Schraefel, M., and Shneiderman, B. From keyword search to exploration: How result visualization aids discovery on the web. Human-Computer Interaction Lab Technical Report HCIL-2008-06, University of Maryland (2008).Google Scholar
- Leide, J., Cole, C., Beheshti, J., Large, A., and Lin, Y. Task-based information retrieval: Structuring undergraduate history essays for better course evaluation using essay-type visualizations. Journal of the American Society for Information Science and Technology 58, 9 (2007), 1227--1241. Google ScholarDigital Library
- Li, L., Chu, W., Langford, J., and Schapire, R. A contextual-bandit approach to personalized news article recommendation. In Proc. of WWW (2010). Google ScholarDigital Library
- Marchionini, G. Exploratory search: from finding to understanding. Commun. ACM 49, 4 (Apr. 2006), 41--46. Google ScholarDigital Library
- Matejka, J., Grossman, T., and Fitzmaurice, G. Citeology: visualizing paper genealogy. In Proceedings of Human Factors in Computing Systems Extended Abstracts (2012), 181--190. Google ScholarDigital Library
- Pandey, S., Agarwal, D., Chakrabarti, D., and Josifovski, V. Bandits for taxonomies: A modelbased approach. In Proc. of ICDM (2007).Google Scholar
- Pu, P., Chen, L., and Hu, R. A user-centric evaluation framework for recommender systems. In Proc. of RecSys (2011), 157--164. Google ScholarDigital Library
- Radlinski, F., Kleinberg, R., and Joachims, T. Learning diverse rankings with multi-armed bandits. In Proc. of ICML (2008), 784--791. Google ScholarDigital Library
- Seo, Y., and Zhang, B. A reinforcement learning agent for personalized information filtering. In Proc. of IUI (2000), 248--251. Google ScholarDigital Library
- Spärck-Jones, K. A statistical interpretation of term specificity and its application in retrieval. Journal of Documentation 28, 1 (1972), 11--21.Google ScholarCross Ref
- Spink, A., Wolfram, D., Jansen, M. B. J., and Saracevic, T. Searching the web: The public and their queries. Journal of the American Society for Information Science and Technology 52 (2001), 226--234. Google ScholarCross Ref
- Teevan, J., Alvarado, C., Ackerman, M. S., and Karger, D. R. The perfect search engine is not enough: a study of orienteering behavior in directed search. In Proc. of SIGCHI (2004), 415--422. Google ScholarDigital Library
- Terveen, L., Hill, W., and Amento, B. Constructing, organizing, and visualizing collections of topically related web resources. ACM Transactions on Computer-Human Interaction (TOCHI) 6, 1 (1999), 67--94. Google ScholarDigital Library
- Tombros, A., and Sanderson, M. Advantages of query biased summaries in information retrieval. In Proc. of SIGIR (1998), 2--10. Google ScholarDigital Library
- White, R. W., Bennett, P. N., and Dumais, S. T. Predicting short-term interests using activity-based search context. In Proc. of CIKM (2010). Google ScholarDigital Library
- Yee, K.-P., Swearingen, K., Li, K., and Hearst, M. Faceted metadata for image search and browsing. In Proc. of SIGCHI (2003), 401--408. Google ScholarDigital Library
- Yue, Y., and Joachims, T. Interactively optimizing information retrieval systems as a dueling bandits problem. In Proc. of ICML (2009). Google ScholarDigital Library
- Zhai, C., and Lafferty, J. A study of smoothing methods for language models applied to information retrieval. ACM Trans. Inf. Syst. 22, 2 (Apr. 2004), 179--214. Google ScholarDigital Library
- Zhang, B., and Seo, Y. Personalized web-document filtering using reinforcement learning. Applied Artificial Intelligence 15 (2001), 665--685.Google ScholarCross Ref
- Zhang, Y., Xu, W., and Callan, J. Exploration and exploitation in adaptive filtering based on bayesian active learning. In Proc. of ICML (2003), 896--903.Google Scholar
Index Terms
- Directing exploratory search: reinforcement learning from user interactions with keywords
Recommendations
Directing exploratory search with interactive intent modeling
CIKM '13: Proceedings of the 22nd ACM international conference on Information & Knowledge ManagementWe introduce interactive intent modeling, where the user directs exploratory search by providing feedback for estimates of search intents. The estimated intents are visualized for interaction on an Intent Radar, a novel visual interface that organizes ...
Content-based book recommending using learning for text categorization
DL '00: Proceedings of the fifth ACM conference on Digital librariesRecommender systems improve access to relevant products and information by making personalized suggestions based on previous examples of a user's likes and dislikes. Most existing recommender systems use collaborative filtering methods that base ...
A Study on User-Controllable Social Exploratory Search
IUI '18: Proceedings of the 23rd International Conference on Intelligent User InterfacesInformation-seeking tasks with learning or investigative purposes are usually referred to as exploratory search. Exploratory search unfolds as a dynamic process where the user, amidst navigation, trial-and-error and on-the-fly selections, gathers and ...
Comments