Supporting multiple information-seeking strategies in a single system framework (abstract only)
Xiaojun Yuan · 2004
This research explores the relationship between information-seeking strategies (ISSs) and information retrieval (IR) system design. When people seek information they engage in a variety of ISSs in order to search for specific items, learn about the contents of the database, evaluate retrieved information, and so on.The theoretical foundations of the work are based on the information-seeking episode model developed by Belkin (1996), and the multi-facet classification scheme of information behaviors proposed by Cool & Belkin (2002).The goal of this research is to construct and evaluate an interactive retrieval system which uses different combinations of IR techniques to support different ISSs. Example IR techniques include comparison using exact and probabilistic matching algorithms; summarization of information objects using titles, snippets or abstracts; visualization techniques such as lists or classified results; and navigation techniques such as scrolling or following links. By designing a retrieval system with diverse strategies in mind, we can adaptively support multiple ISSs, permitting a user to move seamlessly from one strategy to another, choosing instantiations of each support technique tailored to the specific ISS.The research will be conducted in a series of four steps. (1) Develop an object-oriented framework for representing basic IR techniques. (2) Design, implement and evaluate systems which support individual ISSs such as browsing and searching. (3) Specify an interaction structure for guiding and controlling sequences of different supporting techniques.(4) Design, implement, and evaluate a dynamically adaptive system supporting multiple ISSs in comparison to a non-adaptive baseline system.