Research And Application Of Web Mining For Association Rules

With the rapid development of Internet technology, all kinds ofinformation can be acquired through Internet at a very low cost. At the sametime, since the WWW all around the world is linked, how to calculate thedata we can get from it is very difficult.Just for this reason, how to find useful message in Internet has becomea pop research question. Web data mining is an important tool that can beused to find potential useful information in Internet. There are a lot ofmethods in web data mining fileld, and association rules is an importantaspect of web data miningresearch, it is very necessary to study it.The structure of this paper is organized as follows. Firstly, thepurpose of coposing and the main content of this thesis are explained, thebackground and related work of web data mining and mining association rulesare introduced. During the second part, a suggeston for application ofassociation rules is put forward, the bottleneck probleom of Apriorialgorithm useage is analyzed, and three optimized algorithm are described.Finally, a new incremental, updating algorithm for maintaining discoveredassociation rules is presented when the transaction database increases andthe minimum support changes as well. Experimental results have shown thatthis algorithm can be succeccful used in planning and constructing of e-government portal website.