An Evolutionary Approach toward Search Engine Results Clustering

Web Based Information Retrieval (WBIR) is a system for retrieval of information on the internet and known as the basis of the web search engines. This textbook was originally a PhD dissertation with the title "Search Engine Results Clustering using a Modified Imperialistic Competitive Algorithm". In this study a search engine,MISE, was developed based on a new four-layered core structure that is less complicated than the other existing core architectures for search engines such as Google or Yandex. A redefined version of Imperialistic Competitive Algorithm(ICA) was used for large scale data clustering.Five common search engines(Google,Bing, Yandex, Yahoo and AOL) and ten current page ranking algorithms (GPR, WPR, WLRank, TR, TIR, HITS, Clever, JBR, QDR and DRA) were compared to MISE in terms of precision scores.This study had a number of achievements and contributions, including: design of a content-oriented clustering algorithm for search engines; proposed a new core architecture for search engines; and implementation and comparison with current page ranking algorithms.