Web users access paths clustering is important to conduct Web page prediction. In this paper, a novel Web users access paths clustering method is proposed based on possibilistic and fuzzy sets theory. Firstly, a similarity measure method of access paths is proposed based on differences between paths’ factors, such as the length of time spent on visiting a page, the frequency of a page accessed and the order of pages accessed. Furthermore, considering that clusters tend to have vague or imprecise boundaries in the path clustering, a novel uncertain clustering method is proposed based on combining advantages of fuzzy clustering and possibility clustering. A
_cut set is defined here to process the overlapping clusters adaptively. The comparison of experimental results shows that our proposed method is valid and efficient.