MDdata1, MDdata2

MDdata1和MDdata2是由天津大学等机构构建的两个新数据集，用于训练和评估机器学习中的实体识别模型。MDdata1包含15,236条句子，来源于PAKDD和ACL会议的论文实验章节，通过人工标注和数据增强技术构建。MDdata2则包含58,464条句子，用于分析方法的发展。这两个数据集主要用于机器学习和数据挖掘领域的文献分析和算法推荐，旨在通过提取和分析论文中的方法和数据集信息，补充传统文献分析的不足，反映学术发展趋势。

MDdata1 and MDdata2 are two newly developed datasets constructed by Tianjin University and other institutions, intended for training and evaluating entity recognition models in the field of machine learning. MDdata1 comprises 15,236 sentences sourced from the experimental sections of papers published in the proceedings of PAKDD and ACL conferences, and was built through manual annotation and data augmentation techniques. MDdata2 contains 58,464 sentences and is designed for the development of analytical methods. These two datasets are primarily utilized for literature analysis and algorithm recommendation in the domains of machine learning and data mining. Their core purpose is to supplement the deficiencies of traditional literature analysis by extracting and analyzing information regarding methods and datasets in academic papers, thereby reflecting the trends of academic development.