A novel prime numbers based hashing technique for minimizing collisions

Abstract

Searching is a prime operation in computer science and numerous methods has been devised to make it efficient. Hashing is one such searching technique with objective of limiting the searching complexity to O (1) i.e. finding the desired item in one attempt. But achieving complexity of O (1) is quite difficult or usually not possible. This happens because there is no perfect mapping function for insertion and searching; and this imperfection of hashing function results in collisions. The algorithm and technique presented in this article minimizes the number of collisions by removing the problem of clustering. Clustering occurs when the data items congregates in one particular area thus increasing the number of collisions and results in increased number of probes to insert and search an item. During trials runs the proposed algorithm have shown considerable improvements over all major hashing algorithms in terms of performance.