Keyword-Based Information Retrieval System b Utilizing Vector Space Model (VSM)

Abstract

The majority of today’s Information Retrieval tasks are based on two main processes: indexing and searching. Indexes and retrieval are at the core of every modern Information Retrieval (IR) System. Many IR systems are based on vector space model (VSM). This paper is focused attention on the VSM of searching and retrieving on “Natural Disaster” information. In this system, index table is used for data structure and VSM also applied to retrieve information which is user’s desirable information. Cosine similarity method is used to compute the similarity between query vector and document vector. HTML parser is used in our system for parsing the HTML pages. And we also use the term frequency and inverse document frequency (tf-idf) weighting scheme for the document indexing. Query is formulated by using terms and then this query is evaluated by processing the index to retrieve all stored information by an appropriate combination of query words.