| Summary: | 碩士 === 國立雲林科技大學 === 資訊管理系碩士班 === 95 === In this WWW environment with a large number of information, users usually can’t find results that they exactly need when doing information retrieve. For example, the google search engine responses users’ query requirements in a very short time. But it provides too many results and users still need to filter these results by themselves. The main reason for this situation is that search engine not really understands what users’ requirements are. Therefore a large number of researches about information retrieve continually proceeding in order to provide faster and more precise search results in recent years. Besides conditional query, they often provide advanced multiple conditional query in shopping websites to reduce the range of merchandise data. This kind of search process is more complicated and users must know information about merchandise (ex. size, price, brand, performance etc.). It cost many users’ energy on operating. Therefore, comparing similarities between documents can bring users more convenience in particular applications which data properties are similar. Compare with conditional query, an integrated similarity query requirement is more and more important. Contributions of this research are as follows:
1. Constructing a knowledge management platform which is FCA-Based and applied to structured documents. It will explore and manipulate knowledge included in structures documents.
2. Take “Notebook Catalogs Similarity Query System” for example to be a prototype of relative applications development.
3. Combine user feedback mechanism to raise users’ satisfactions. User feedback mechanism makes the system trainable and react influences caused by hidden information.
|
|---|
