Mining Preferred Browsing Paths Based on Adjacency List Models

Mining preferred browsing paths from Web logs can help improve the structure of Web sites and increase the click-stream. In this paper, an algorithm of mining preferred browsing paths based on adjacency list model is proposed. The Web pages are denoted by vertexes of a directed graph and the topology structure of Web sites is represented by adjacency list models based on the support-preference concept. Moreover, preferred browsing paths are discovered by using depth first search method. Experimental results show that the algorithm can maintain accuracy but have better efficiency than the NPPMA algorithm does.