Galois Lattice for Web Sequential Patterns Mining

Along with the rapid development of the Internet, Web has become most importmant resource of information. It is getting more difficult to fast and accurately find useful information from the gigantic amount of data provided by Web. The Web usage mining is the process inferring valuable knowledge by analyzing users´ access information, which is of great significance for better managing Web sites and further developing good electronic education. In this paper we applied Galois lattice to mine Web sequential access patterns by representing the paths traversed using graphs. An incremental and efficient mining algorithm was put forward for increasing specific day´ Web log and changing transaction base. Experimental results on real-life data sets demonstrated the effectiveness of the novel algorithm.