A Study of an Efficient Approach to Mining Frequent Rooted Embedded Ordered Subtrees

• Main Authors: Yu-ling Liu, 劉毓玲
• Other Authors: Chien-I Lee
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/57461925906254255792

碩士 === 國立臺南大學 === 數位學習科技學系 === 95 === In recent years, many researches put their attention on frequent subtree mining. The application of frequent subtree minining includes the web usage mining, biological information, XML structure, and so on. One important issue of frequent embedded subtree mining is to reduce the computational time. In this paper, we proposed a novel method, called TCMR, to efficiently mining frequent embedded subtrees. First, TCMR combines the concept of m-complete tree and the depth first traversal to show the relationship between nodes in the tree. Furthermore, TCMR extends the idea of embedded list to record instances, so it can reduce the execution time and memory requirement for the recording of the instances. Finally, TCMR effectively controls the scope of join between nodes and therefore reduces the computational time. The experimental evaluation shows that the effectiveness of our algorithm is superior to PrefixTreeESpan.