Information Granulation for the Design of Granular Information Retrieval Systems

With the explosive growth of the amount of information stored on computer networks such as theInternet, it is increasingly more difficult for information seekers to retrieve relevant information.Traditional document ranking functions employed by Internet search engines can be enhanced toimprove the effectiveness of information retrieval (IR). This paper illustrates the design anddevelopment of a granular IR system to facilitate domain specific search. In particular, a novelcomputational model is designed to rank documents according the searchers’ specific granularityrequirements. The initial experiments confirm that our granular IR system outperforms a classicalvector-based IR system. In addition, user-based evaluations also demonstrate that our granular IRsystem is effective when compared with a well-known Internet search engine. Our research workopens the door to the design and development of the next generation of Internet search engines toalleviate the problem of information overload.