CSMT数据挑战数据集

CSMT数据挑战数据集是由北京邮电大学创建，用于识别计算机生成的旋律。该数据集包含两个部分：开发数据集和评估数据集，分别包含6000和4000条记录。开发数据集仅包含计算机生成的旋律，而评估数据集则混合了计算机生成和人类创作的旋律。数据集的创建旨在促进计算机生成旋律的识别方法的发展，并辅助生成音乐系统的创建。该数据集的应用领域主要集中在法律实践中，特别是在版权保护方面，帮助识别音乐来源，解决版权侵权问题。

The CSMT Data Challenge Dataset was developed by Beijing University of Posts and Telecommunications for the task of recognizing computer-generated melodies. It consists of two subsets: the development dataset and the evaluation dataset, which contain 6000 and 4000 records respectively. The development dataset exclusively features computer-generated melodies, whereas the evaluation dataset comprises a mix of computer-generated and human-composed melodies. This dataset is designed to promote the advancement of methods for identifying computer-generated melodies and support the development of music generation systems. Its primary application domains fall within legal practice, specifically copyright protection, where it aids in identifying music sources and resolving copyright infringement issues.