Study Of Scheduling Strategy For Meta Search Engine

With the explosive increase of network information, it becomes more and more difficult for people to look for what they want. To get more full scale, accurate search results, multi-search engines should be used and meta search engines occur. Scheduling strategy is a key factor that can affect the query capability of meta search engines and it is also an important research field. It is to help choose proper group of search engines for a query to get more accurate result for the users with less resource. The basic conceptions of meta search engines are introduced which discuss the factors that affect the efficiency of meta search engines and the scheduling strategy of meta search engines is especially studied in this paper. Scheduling strategies of meta search engines that have appeared in and out of China were studied, also their advantages and shortcomings were analyzed. Then it is detected that some scheduling strategies don't perform well when searching items which are new or less used, and some don't evaluate member search engines definitely. Therefore, a new method is proposed to evaluate the query capability of search engines based on category which uses recall, precision and average response time. Combining this evaluating method and the method based on former experiences, a new scheduling strategy for meta search engines based on study is brought up. A test system was designed and experiments were taken to test the establishment of scheduling strategy, the precision and the query capability for items which are used less. The results show that the test system promotes the precision effectively and improves the query quality of items which are used less.