The Effect of Local Website Search

Global search engine is a huge but it can´t replace local search engine. The local search engines help the end-user to retrieve easily what they want and to give accurate results than global search engines do. The end-user can also use the global search engines to augment the search performed by a local search engine, restricted to a single site. However, a single website is different from the whole web in link structure, access pattern, and possibility hit on page. To improve the performance of search sites, many researches have tried to combine original search methods with new ones such as: the PageRank, LPageRank, predictive access pattern,... This article would like to introduce a hybrid model to improve local website searched by using web server logs. This model combines capability of ranking structure link of PageRank and recommending access pattern of neural network to improve the result of search engine. It also present the way to combine search engine with PageRank, web log and neural network.To experiment with the new model, the authors have used this model for Music Machine Website with their web logs.