Meta-search Engine Key Technologies

As a highlight in the phylogeny of Internet and Web, search engine becomes more and more important in the development of WWW by right of the advantages of itself, and also it becomes the hotspot of information retrieval (IR). Especially in recent several years, the huge, heterogeneous and dynamic web information has been in an embarrassed position of "rich information, poor knowledge", which makes it essential to find an efficient and effective method to visit Web information. Also, with the continuous upgrade of the huge , heterogeneous and dynamic nature of Web information, independent search engines are incapable of processing the Web search problem. All these problems make the way for Meta search engine.Meta search engine is a software system of Web IR. It transmits the query to several independent search engines, and then merges the results collection returned from independent search engines with its own special technology, and finally displays the results to users.This thesis pays attention on the necessity and current limitations of Meta search engine based on the overview of independent search engines and Meta search engines. In the end, the two key problems of Meta search engine: search engine selection problem and technology of result merging are discussed. Detailed as following:(1) The overview of the independent search engines and the Meta search engines are given, and the deficiency of current research work is brought in.(2) Introduce and analyze the current main methods of search engine selection problem, and then a new method that based on artificial neural networks is proposed.(3) Introduce and analyze the current main methods of removing repeated records and technology of result merging, and then a new methods is proposed.