Information Retrieval Research Group
Our main research areas are:
- cognitive and user-centric methods for the design and evaluation of IR systems;
- the development of new statistical IR techniques;
- multimedia browsing and retrieval; and
- personal information management and retrieval.
We have a strong interest in cognitive and behavioural approaches, including extensive studies of information-seeking behaviour, and of the design of user interfaces for interactive text searching/multilingual retrieval. We are also investigating interfaces for multimedia retrieval and how people access and retrieve personal data such as emails, desktop files and meeting records. Research in the statistical IR area includes: the use of linguistic methods for query expansion, cross-language IR, information extraction, multi-media metasearch, geographic IR, peer-to-peer search, retrieval of cultural heritage materials, and speech retrieval. In the area of multimedia retrieval we are researching novel techniques for searching and browsing image databases, as well as speech and video corpora. Finally we are interested in investigating new methods and measures for evaluating interactive information retrieval systems. In particular we focus on usability issues and how empirical studies can better support/influence interactive system design.