A new context-based information retrieval system
Vivek Chanana, Athula Ginige, San Murugesan · International Conference on Artificial Intelligence · 2004
Abstract : - We present a new context-based information retrieval (IR) system where user could indicate the context in which he/she is seeking information. We use a novel way of representing context based on the type of information. The context defined in this manner is most intuitive to the user and originates from his/her information need in a problem-solving situation. This new context-based information retrieval system overcomes major limitations of the keyword based IR systems currently in use and provides improved performance. Key-Words: - Information retrieval, context, retrieval effectiveness, information type, information need, user interface. 1 Introduction The subject of context has received a great deal of attention in the information retrieval literature over the past decade, primarily in studies of information seeking and information retrieval interactions. The information search environment has changed from a physical traditional library to online on the Web, and then recently to more ubiquitous, mobile environments. Researchers are presented with new challenges in dealing with information retrieval in this new environment. One of the failings of most information retrieval models has been attributed to the isolation of queries from the context in which they occur [2], or the context of the information need. Where a search intermediary, such as a librarian, is available, the intermediary is able to establish the context of a user’s search for information, and hence advise and guide a user in information retrieval. However, such assistance or interaction with a human intermediary, nor provision for seeking information in a particular context, is available in Web search engines and other automated information retrieval systems. User queries without context can be vague and result in the retrieval of many non-relevant documents. This is a classic problem in information retrieval - the retrieval of documents that have the right terms, but in the wrong context [15]. This is a major concern that is becoming increasingly important in the Web environment, where user queries are typically short and system fails to disambiguate them adequately, resulting in numerous irrelevant information. Specific queries with proper context are more likely to reduce some non-relevant documents. Context can reduce ambiguity by associating meanings with request terms and thus limit the scope of the possible interpretations of query terms. The context-based information retrieval framework provides a foundation of effective information retrieval system.