An improved Approach for Document Retrieval Using Suffix Trees

Abstract

Huge collection of documents is available at few mouse clicks. The current World Wide Web is a web of pages. Users have to guess possible keywords that might lead through search engines to the pages that contain information of interest and browse hundreds or even thousands of the returned pages in order to obtain what they want. In our work we build a generalized suffix tree for our documents and propose a search technique for retrieving documents based on a sort of phrase called word sequences. Our proposed method efficiently searches for a given phrase (with missing or additional words in between) with better performance. Keywords-Document retrieval; Frequent Word Sequences; Suffix tree; Traversal technique. I. INTRODUCTION With the growth of web, hundreds of millions of people engage in information retrieval process every day when they use web search engine or search their emails. IR is fast becoming the dominant form of information access, overtaking traditional database style searching. IR process begins when user enters a query like search strings in web search engines, phrases etc. to identify the related documents or URLs. Now almost all the documents have electronic copies. With the development of WWW it is an efficient technique to retrieve the documents using the web search engines based on a query. But this should not be time consuming. That is the reason precision of the retrieval of related documents for a given query is vital for the search engine. Cluster based information retrieval techniques also exist (11). The next section deals with the Information Retrieval and its related work on text documents. Section 3 describes Suffix Tree. Section 4 deals with building generalized suffix tree. Section 5 explains traversal technique Algorithm used for quick retrieval of documents. Section 6 shows the experiment