Farid Soleymani Sabzchi - Department of Computer Engineering, Zanjan Branch, Islamic Azad University Zanjan, Iran
Mehdi Afzali - Department of Computer Engineering, Zanjan Branch, Islamic Azad University Zanjan, Iran

Developing word of the Web, increasing the content information and requirements of user’s Web site has been changed. Therefore, Web needs a dynamic and an accurate algorithm to recognizeuser’s requirements to suggest new patterns. There are many algorithms proposed for discovering frequent patterns. Miningfrequent patterns is one of the fundamental and essential operations in many data mining application such as discoveringassociation rules. In various applications, database frequentlychanges by inserting, deleting and modifying transaction. The proposed algorithm has the potential to apply these four factorsto modify database in the path tree by incremental mining. This algorithm has been compared to similar algorithms such asCATS-tree, AFPIM, CAN-tree and CP-tree. Suggested algorithm has lower time complexity and higher speed in compare withother algorithms. To describe this algorithm, an illustrative example is presented. Obtained results show that extracted patterns by this method will specify degree of user’s interest tothe pages more accurately and also the steps of sorting branches after applying any change in the tree has the lowest executive order in compare with other algorithms

Incremental Mining, Path Tree, Frequent Patterns, association rules