Indexing web access-logs for pattern queries

In this paper, we develop a new indexing method for large web access-logs. We are concerned with pattern queries, which advocate the search for access sequences that contain certain query patterns. This kind of queries find applications in processing web-log mining results (e.g., finding typical/atypical access-sequences). The proposed method focuses on scalability to web-logs' sizes. For this reason, we examine the gains due to signature-trees, which can further improve the scalability to very large web-logs. Experimental results illustrate the superiority of the proposed method.