Abstract: Multi-label learning, which considers the case of an object related to multiple labels, attracts much attention in recent years. Multi-label learning research aims to improve the performance of multi-label learning algorithms by reducing the complexity of the feature space and the label space. This paper systematically analyses the developments in multi-label learning research from four aspects including multi-label classification, label ranking, multi-label dimension reduction and label correlation and also points out the existing problems in the multi-label learning research. Finally, it summarizes several valuable research directions, which provides reference for the further research in this field.

Key words: multi-label learning, classification, label correlation, dimension reduction

摘要： 多标记学习考虑一个对象与多个类别标记相关联的情况，是当前国际机器学习领域研究的热点问题之一。多标记学习的研究主要围绕降低特征空间和标记空间的复杂性，提高多标记学习算法的精度而展开。针对这一特点，从多标记分类、标记排序、多标记维度约简和标记相关性分析四个方面，对多标记学习的研究进展进行了归纳与阐述，分析了当前多标记学习存在的问题。最后指出了目前多标记学习若干发展方向，为该领域的进一步研究提供参考。

关键词: 多标记学习, 分类, 标记相关性, 维度约简