Java方法与描述数据集

本数据集由圣母大学计算机科学与工程系创建，包含超过2.1百万对Java方法及其自然语言描述，源自2.8万个Java项目。数据集旨在支持源代码摘要自动生成的研究，通过清理和标记化处理，确保数据质量。该数据集适用于解决软件文档自动化生成的问题，特别是在JavaDoc方法描述的自动生成方面。

This dataset was created by the Department of Computer Science and Engineering at the University of Notre Dame. It comprises over 2.1 million pairs of Java methods and their natural language descriptions, sourced from 28,000 Java projects. This dataset aims to support research on automatic source code summarization, with data quality ensured through cleaning and tokenization processes. It is applicable to addressing the problem of automated software documentation generation, particularly for the automatic generation of JavaDoc method descriptions.