点击次数：

论文类型：会议论文

发表时间：2011-09-16

收录刊物：EI、CPCI-S、Scopus

卷号：433-440

页面范围：2250-2256

关键字：P2P; content-based; music retrieval; index

摘要：In view of existing situation that large collections of music data are shared by numerous users in Peer-to-Peer network, people raise higher demands for content-based music information retrieval. For more efficient searching, this paper mainly involves the following three points: Firstly, filter out the repeating patterns in original piece of music and extract the key information, which can reduce the space occupied by index structure. Secondly, we propose a music index structure based on ordered orthogonal list to store music features in a nested sparse matrix. Finally, a related search table is proposed according to user's search history to store the global connectivity of the network. Experimental results show that the proposed model can reduce the traffic cost and average response time effectively and have high retrieval accuracy.