Locality sensitive hashing based searching scheme for a massive database

Abstract

The rapid growth of information nowadays makes efficient information searching increasingly important for a massive database with tremendous volume of information. Traditional methods either rely on linear searching or depend on a tree structure. These methods search information in the entire database and compare a query with the records in the database during the searching process, which lead to inefficiency. This paper presents a locality sensitive hashing based searching scheme (LSS) to achieve highly efficient information searching in a massive database. LSS classifies information based on their similarities to facilitate fast information location. Based on the study and analysis of LSS, an improved scheme is further proposed to enhance the searching efficiency. Simulation results demonstrate the efficiency and effectiveness of the LSS schemes in searching information. They yield significant improvements over the efficiency of traditional methods. In addition, they guarantee successful location of the queried records.