Optimization of Ranking Algorithm in the Information Retrieval Process Using Evolutionary Computation

Abstract

In today’s era, with the increase of digital information at a rapid pace, there has been a tremendous change in the World Wide Web which urges the researchers to find different ways to manage this enormous flow of information and help in satisfying the need of the user. It has been observed that the core component of any search engine is the ranking framework that helps to improve the result quality by ranking the web pages based on user queries. A good ranking model should incorporate diverse measures that ameliorate the result quality and should not stick to a single measure. In the Information Retrieval process, the process of ranking can be enhanced by incorporating process of content similarity and link analysis both. In coming years, it has been observed that various nature-inspired ranking algorithm can be used which integrate different measures and help to optimize the quality of the result obtained. Search techniques that are inspired by nature are referred as evolutionary algorithms and are regarded as a population-based stochastic algorithm. The development of various approaches helps in retrieving the information efficiently and in this regard nature-inspired algorithms like Swarm Intelligence, Genetic Algorithm play a major role. In ranking, we associate web pages stored in the repository in order of preference, so that the pages which have high similarity index according to the user query are ranked in higher-order in comparison to other web pages. For any search engine, ranking algorithm is considered as an indispensable part which overall effect the processing result. In minimum possible time, the most beneficial results should be displayed by the ranking algorithm. The intention of my research work is to optimize the ranking algorithm using different measures that will facilitate the Information Retrieval process thereby help the user in retrieving the relevant information in minimum time duration.