Multi-label Classification Trending Challenges and Approaches

Multi-label classification (MLC) has caught the attention of researchers in various domains. MLC is a classification which assigns multiple labels to a single instance. MLC aims to train the classifier for modern applications such as sentiment classification, news classification, and text classification. MLC problem can be solved by either converting into a single-label problem or by extending machine learning methods for solving it. In this paper, the challenges faced during training the classifier which includes label space dimensionality, label drifting, and incomplete labeling are considered for review. This paper also shows the newly emerged data analysis methods for multi-label data.