Information Retrieval System Using BM25, Pivoted Normalization and Combsum methods

Abstract

Nowadays, a lot of information is available for someone who wants to find documents on Google or in a system or in a digital library. Information retrieval is important as a way that can solve problem by giving the information that seems to be related to documents. Retrieving information is difficult and time consuming for searching a variety and large number of documents on the digital library. The proposed system is intended to implement an effective searching system that intends to implement effective search system for digital library. BM25 and Pivoted Normalization are the best retrieval models for information retrieval system. The CombSUM is combining these two methods to get more relevant documents and to give better output results. The proposed system can help the user to get all relevant documents (conference papers) according to the given query. When a user enters the query, the most relevant documents are ranked by using BM25, Pivoted Normalization Method and CombSUM.