Research On Key Technologies Of Theme-based Meta Search Engines

In recent years, the World Wide Web is developing at an amazing speed,the amount of information has increased dramatically. How the user can obtain the most useful information in the shortest time become a primary problem. In order to solve these problems, meta-search engine has been developed. Its working principle is to integrate many independent search engines, gather their returning results by certain dispatching strategy and the merging algorithm of results. This method not only can expand the information coverage, but also can choose suitable results through the user preferences.The paper describes the current limitations of search engine development and analyses the purpose to study the meta-search engine. It also introduces the basic concept of meta-search engine and the working principle. Then it describes the two key techniques of the meta-search engine: member search engine dispatching strategy and results merging algorithm, and analyses the current research at home and abroad. This paper proposes a meta-search engine model based on the theme, compared with traditional search engine, it introduces the user focus model, the user input analysis algorithm, the dispatching strategy based on the pseudo crawler and database. For member search engine dispatching strategy, it is based on the analysis of the intention of users'queries, the similarity of member search engine database, the theme and the users'focus. For the results merging algorithm, the results are sorted by their position in the member search engine, users's focus on member search engines and the relevance to the subject category. To remove the repetitive results, here it uses a method based on the URL, title and abstract.Finally, this paper analyzes the algorithm experimental results and gives the further steps of the research.