Towards an Adaptive Exploratory Search Framework

Topic Description

Conventionally, an information retrieval (IR) system (or a search engine) returns a ranked list of documents in response to a given user query. However, it is insufficient to tackle a common but under-supported information seeking activity, which is interactive and exploratory in nature and involves learning through the interactive information exploration process (referred to as exploratory search (ES)), often comprising multiple search and browsing sessions over a domain. Typical examples are a researcher browsing literature to build a understanding of a new research domain; or a journalist tracing a series of news events discussed on social media sites and the users's comments/opinions. Nevertheless, how such ES scenarios can be effectively modelled and deployed remains unsolved. ES, for both users and systems, is an interactive and continuous adaptation process. Users’ information needs and retrieval contexts, and even relevance judgments evolve over time, within and beyond a particular information seeking episode. The system and users need to continuously adapt to each other: i.e., the system learn from the users’ interactions and search behaviours; and the users are guided through (yet have control of) the information space step by step according to the exploration paths dynamically suggested by the system. This project aims to develop, deliver and evaluate a next generation information retrieval paradigm by developing novel computational methods and user interaction support to facilitate interactive exploration over online information spaces, which are first pre-processed into fine-grained categorical knowledge structures, e.g., in the form of entity association networks. This project will be in collaboration with University College London.

Skills Required:

Applicants must have a high quality Honours Degree (preferably First Class) or a Master qualification (preferably with distinction) in a relevant discipline; Knowledge of probability theory and statistics; programming skills; Research experience and publications in information retrieval, natal language processing and/or human computer interaction would be desirable.

Background Reading:

Qu, Y., and Furnas, G. (2007). Model-driven Formative Evaluation of Exploratory Search: A study under a sensemaking framework. IP&M, 44(2), 534-555.
White, R.W. and Roth, R.A. (2009) Exploratory Search: Beyond the query-response Paradigm.
Lau, R.Y.K, Song, D., Li, Y., Cheung, T. and Hao, J. (2009). Towards A Fuzzy Domain Ontology Extraction Method for Adaptive e-Learning. IEEE TKDE, 21(6), 800-813.
Liu, H., Mulholland, P., Song, D., Uren, V. and Rueger, S. (2010). Applying Information Foraging Theory for Understanding User Interactions with Content-based Image Retrieval. IIiX10, 135-144.
Fu, W., Kannampallil, T.G., and Kang, R. (2010). Facilitating Exploratory Search by Model-based Navigational Cues. IUI10,199-208.
Cervino Beresi, U., Kim, Y., Song, D. and Ruthven, I. (2011). Why Did You Pick That? Capturing reasons for assigning value in exploratory search. IJDL, 11(2), 59-74.
Clark, M., Kim, Y., Kruschwitz, U., Song, D., Albakour, D., Dignum, S., Cervino Beresi, U., Fasli, M. and De Roeck, A. (2012). Automatically Structuring Domain Knowledge from Text: a review of current research. Information Processing and Management (IP&M). Vol. 48(3), pp. 552-568.
Adeyanju, I., Song, D., De Roeck, A., Albakour, M-D., Kruschwitz, U. and Fasli, M. (2012). Adaptation of the Concept Hierarchy Model with Search Logs for Query Recommendation on Intranets. Proceedings of ACM SIGIR2012, pp. 5-14.
Tan, B., Lv, Y., Zhai, C. (2012). Mining long-lasting exploratory user interests from search history. CIKM 2012: 1477-1481.
Karimzadehgan, M., Zhai, C. (2013). A learning approach to optimizing exploration-exploitation tradeoff in relevance feedback. Information Retrieval. 16(3): 307-330.



Report an error on this page