Mining Web navigation patterns with a path traversal graph

With the expansion of e-commerce and mobile-based commerce, the role of web user on World Wide Web has become pivotal enough to warrant studies to further understand the user’s intent, navigation patterns on websites and usage needs. Using web logs on the servers hosting websites, site owners and in turn companies, can extract information to better understand and predict user’s needs, tailoring their sites to meet such needs. The former mining algorithms do not provide a clear picture of the intentions of the visitors and suffer from drawback of either repetitive database scan or high memory load. This paper uses the concept of throughout-surfing patterns(TSPs) and proposes an efficient algorithm for mining the patterns, that effectively predict and display the trends toward the next visited Web pages in a browsing session with a view to better understand the purposes of website visitors. It also uses a compact graph structure, termed a path traversal graph, to record information about the navigation paths of website visitors, required for mining TSPs. In addition, it proposes a new algorithm for graph traversal based on the prior graph structure to discover the TSPs. The experimental results show the proposed algorithm is highly efficient to discover TSPs, by improving the accuracy, reducing the execution time and memory requirements with a single scan on the database & avoiding generation of candidate sequence as like apriori. Keywords— Web Log Mining, Path traversal graph, Throughout-surfing pattern, Browsing behavior, web Traversal pattern ——————————  ——————————