The Research Of Website Structure Optimization Technology Based On Web Mining

Benefited from the innovation of Web technology, especially the rapid development of theInternet of things, the enterprises get the ability to recognize and perceive the physical world bysensors, RFID and other intelligent device, relying on the communication network, the enterprisesrealized the information interaction between human and things, the seamless connection betweenthe information of production and life，which have integrated effectively the Infranet, Intranet,Internet, because of this, information on Web become more and more rich, which brings thegreater convenience to people, but meanwhile increases the user’s difficulty of getting theinformation which they want from the mass massive data, in the face numerous links, visitorsoften click the pages which are not relevant to the target page. On the one hand, this can increasedata flow burden of Web server and network, cause easily the network congestion, lead to thereduction of access speed of network, which further influence the efficiency of informationsearching. On the other hand, it can block in the enterprises electronic business activities.At present, solution for such problems is the Web data mining techniques, one of mosteffective which is Website structure optimization, which finds out the inadequacies of the existingsite structure by using data mining techniques and make improvements to optimize the structure ofthe site. Among the technologies in a variety of Website structure optimization, site structureoptimization technology based on Web usage mining is an important research area. Web usagemining techniques can take advantage of access logs to find the user’s access mode, and throughanalysis and research Web log records in the law to guide the conduct of e-commerce Websitestructure optimization.This paper systematically elaborated the whole process of data mining, Web data mining andWeb logs mining, here is the main achievement of the work:(1) The improved algorithm of finding the possible expected location is proposed afterstudying the process of optimizing Website structure by Web log mining. The new algorithmmainly optimizes the algorithm at the point of the earlier and the less backtracks. The analysisindicates that new algorithm can find expected pages effectively, thus can achieve the target ofadjustment and reorganization of Website.(2) The method of data preprocessing in Web log mining for Distributed websites is present.the process of data consolidation is added before data clean-up, logs merge algorithm is also present, which takes into account fully the influence on the log merger because of server clocksynchronization and time zone differences, so we add a page filter section between data cleaningand user identification. Improved frame structure page filtering algorithm, in order to eliminate theimpact of the Framework page on Web log mining algorithms.(3) Designing the site structure optimization tool for college site using the above algorithm.By this tool, user access logs and site structure data can be collected, and then be cleaned up andconversion. The site structure optimization tools use these data to analyze the access of the sitepages, and adjust the site link topology according to these results in order to provide users withconvenient and user-friendly links, optimize the structure of the site.