A personalized query expansion approach using context

When humans communicate, queries made by one human do not completely describe the requirements of the human most of the times. Still the other human involved in the communication can understand the messages better, if they share the same context and therefore know the preferences of each other. Similar to human communication, if an automated system could keep or capture the contextual information related to a user and the query, then it could use this information to process the query, resulting more useful answers. This paper is based on a research related to these factors, understanding user request better by identifying domain and task of user query and expanding user query using contextual information such as domain, user and task-specific preferences. The main steps required for this query expansion, and a domain identification algorithm to identify the query domains were proposed in the research. Sets of rules that could be used to identify the tasks and the relevant user preferences are defined in this paper. The algorithm and the rules sets were implemented and the results of the evaluation are also presented in the paper.