Comparative analysis of K-Means with other clustering algorithms to improve search result

The paper identifies the scope of improvement for the search result of a web site. The study includes some commonly used clustering algorithms to identify the usage of clustering approach for improving web elements analysis, in various ways. As the Search result option is extensively used at almost every web site, the main focus is to optimize search result of a web site using clustering approach. Sementic web using the concept of ontology is included, to retrieve more relevant and meaning full serach results. Some most commomly used algorithms are experimented using web data, and it is observed that K-Means clustering algorithm gives best result in term of accuracy and speed. Thus the proposed hybrid model will be using K-Means and Genetic algorithm to overcome the drawbacks of K-Means. The evaluation parameters; accuracy in terms of objects placement in correct cluster, relevancy, speed and user satisfaction are the main parameters considered for the study.