BEST SEARCH AND RETRIEVAL PERFORMANCE EVALUATION WITH LUCENE INDEXING

Abstract

Transaction processing in organizations commonly use relational database, but big part of database operations need select operation. As data increases with few gigabytes selection operation required more time to process whole transaction. There is one approach get frequently used to build indexes in database on columns which is selection. If number of table's (which is general case)get used then selection takes more time. Another approach we can use which is searching framework for searching records. Apache Lucene is a high performance full featured text search engine library written entirely in java which allows simultaneous update and searching. In this paper, we are using these capabilities of Lucene to improve performance of data retrieval. We are using Lucene indexing types to reduce load on database for selection operations. So leveraging advantages of Lucene search for our need, we propose and implement search history module in Lucene as a plugin so that it can be added at any stage in search framework. So that the search history maintenance for each user will be easy as well as can give good performance for large number of users. This paper surveys Lucene indexing and searching.