点击次数：

论文类型：会议论文

发表时间：2010-01-01

收录刊物：EI、Scopus

卷号：1

页面范围：448-451

摘要：A model of XML document is extended by considering both path and frequency information, namely the frequency-path model. Based on this model, a structural similarity calculation algorithm with position and frequency weight by longest common subsequence (PFWLCS) is proposed, which is fast and has high precision. Furthermore the selection of the position and frequency factors are discussed in depth. Experiments show that the PFWLCS has higher recall ratio and accuracy than existing similarity calculation methods, especially on XML with different Structures. ? 2010 IEEE.