Rough set based concept extraction paradigm for document ranking

Abstract

On the World Wide Web, Open domain Question Answering System is one of the emerging information retrieval systems which are becoming popular day by day to get succinct relevant answers in response of users' questions. In this paper, we are addressing rough set based method for document ranking which is one of the major tasks in the representation of retrieved results and directly contributes towards accuracy of a retrieval system. Rough sets are widely used for document categorization, vocabulary reduction, and other information retrieval problems. We are proposing a computationally efficient rough set based method for ranking of the documents. The distinctive point of the proposed algorithm is to give more emphasis on presence and position of the concept combination instead of term frequencies. We have experimented over a set of standard questions collected from TREC, Wordbook, WorldFactBook using Google and our proposed method. We found 16% improvement in document ranking performance. Further, we have compared our method with online Question Answering System AnswerBus and observed 38% improvement in ranking relevant documents on top ranks. We conducted more experiments to judge the effectiveness of the information retrieval system and found satisfactory performance results. © Springer-Verlag Berlin Heidelberg 2010.