Information Retrieval System with Pseudo Relevance Feedback using Rocchio algorithm

Abstract

Information retrieval is concerned with organizing and retrieving relevant documents from a set of unstructured collections of documents.In most collections, the same concept may be referred to using different word.This issue is known as synonymy which has an impact on the recall of most information retrieval systems.The problem can be eliminated by means of relevance feedback where user judges the result of the initial query, and system refines the query vector to improve the recall.But in normal relevance feedback, users have to give manual feedback which is inflexible, so the proposed system uses pseudo relevance feedback where top-k retrieved documents are assumed to be relevance and query is rewrite using Rocchio algorithm.Finally the system analyzes with precision and recall of information retrieval results