Music plagiarism detection using melody databases

Abstract

This paper addresses the development of a system that detects plagiarism based on similar melody searching. Similar melody searching is to find the melodies similar to a given query melody from a music database. For this purpose, we propose a novel similarity model that supports alignment as well as shifting. Also, we suggest a method for indexing the features extracted from every melody, and a method for processing plagiarism detection by using the index. With our plagiarism detection system, composers can easily search for the melodies similar to their ones from music databases. Through performance evaluation via a series of experiments, we show the effectiveness of our approach. The results reveal that our approach outperforms the sequential-scan-based one in speed up to around 31 times. © Springer-Verlag Berlin Heidelberg 2005.