Abstract: Frequent sequence mining has been a topic of research interest in recent years. Mining of web log sequence with forward references is an application of sequence mining and useful with many applications that include web recommendation, caching, pre-fetching etc. In this study, a fast algorithm presented that aims at mining prefix based maximal frequent contiguous sequences with priority in web logs efficiently. To enable fast mining, the algorithm first generates maximal potential patterns and discovers frequent patterns from them without candidate generation. Experimental results show that the proposed algorithm can perform better than existing approaches.
M. Thilagu and R. Nadarajan, 2010. Discovery of Maximal Contiguous Sequence Patterns with Priority in Web Logs. Asian Journal of Information Technology, 9: 238-242.