Summary: | 碩士 === 逢甲大學 === 資訊工程所 === 90 === With the huge amount of information available on the World Wide Web, Web servers provide a fertile ground for information searches. Although numerous searching technologies have been developed for Web searches, there still has a lot of space for improvement. In this thesis we present a new ranking algorithm and an intelligent Web searching system using data mining techniques to search and analyze Web documents in a more flexible and effective way. Our method employs the characteristics of Web documents to extract, find, and rank data in a meaningful manner. We utilize hyperlink structures along with the content of Web documents intelligently to rank the retrieved results. It can solve the ranking problems of existing algorithms such as multi-frame Web documents and unrelated linking documents. In addition, we use domain specific ontologies to improve our query process and to rank retrieved Web documents with better semantic notion. We also provide more friendly and easy to use interfaces.
Two data mining techniques, association rules mining and clustering are employed online for users to explore and browse the retrieved documents conveniently. We use association rules mining to find the patterns of maximal keyword sets, which represent the main characteristics of the retrieved documents. For subsequent queries, these keywords become recommended sets of query terms for the specific users. Clustering is used to group retrieved documents into distinct clusters that can help users make their decisions easier. We have developed the proposed system and tested it with both Chinese and English Web documents from the Web site of Feng Chia University. The result shows that it is indeed a flexible and effective Web searching system.
|