Obtaining the Name aliases from the Web, using them to Cluster Text Documents with Cuckoo Algorithm and Comparing Results with K-Means Algorithm

Abstract

There is an increase in the searching where name aliases are concerned. Approximately 30 percent of searches are based on aliases; hence it becomes important to obtain correct aliases. Lexical pattern based method is used to obtain the aliases of any personal or place from the web The aliases obtained are ranked and filtered based on the co-occurrence frequency and web dice methods These final aliases are then used to cluster the text documents present in a huge database. To get the best cluster cuckoo method of clustering is used. This method is based on the reproduction system of the cuckoo bird. According to the studies this clustering method when used with levy flight concept gives the best results when huge data is concern and also outperforms particle swarm optimization algorithm and genetic algorithm. The result will be compared with the result of kmeans clustering method.