Music Information Retrieval Schemes in Peer-to-Peer Environments

Abstract

Peer-to-peer systems are useful tools for music applications. Existing peer-to-peer systems support the storing and sharing of music data, however they cannot effectively and efficiently support the content-based information retrieval and the cooperation of musicians. This study focuses on the methods of content-based music information retrieval in peer-to-peer environments. Four music information retrieval schemes are evaluated in detail on communication cost, retrieval time, update complexity and robustness. Peers-peers-coordinator scheme is found to be the best one from theoretical analysis and simulated experiments. Also, an algorithm is designed for an implementation of the peers-peers-coordinator scheme and a simple but effective method is brought forward to filter out the replica in the final results. The results of simulated experiments show the efficiency of the algorithm.