Research Of Web Structure Model Based On Granular Computing

With the rapid development of Web technology, it is becoming increasingly urgent that how to deal with relational questions of Web efficiently. Under this background, a new research direction which was named Web Intelligence (WI) has appeared, and it becomes a hot issue of application in the field of Web theory in recent years. WI is a product which mixing the artificial intelligence and advanced information technology under the new environment of Web and Internet. This dissertation gives a research of Granulation Calculation theory on Web Mining technology, which is the core technology of WI, and brings in the idea of Information Granulation to Web structure. A new Web structure of granularity model is presented, which can solve the improving and optimizing problems of Web organization structure.The main contents include:Firstly, the paper concentrates on studying the development of Web Intelligence technology, describes the organizational structure of Web, discusses the Web mining technology widely, elaborates the theory of Granular Computing(GrC), and research its core idea, especially, introduces the Theory of Rough Set and Theory of Quotient Space.Secondly, the conception of Abstract Granule is outlined. A new granularity model of Web structure is presented basing on the theory of Granular Computing, and a conception of Information Granulation is presented basing on the organizational structure of website. The paper defines some relative impact factors about Granule, such as, Webpage Structure Granularity, Webpage Structure Similarity, Webpage Significancy, Webpage Relational Granule, User Similar Granule, etc. . Finally, some examples are given to calculate this basing on a actual part of website structure.Thirdly, with studying the technology of optimizing website structure, a new heuristic method of calculating information granules is presented basing on the granularity model of Web structure, which can generate efficiently relational rules that can adjust and improve the organizational structure of website, and reach the ultimate purpose of optimizing Web structure. Finally, there is a experimental analysis of the method.