Multi-Label Classification Method Based on Self-Attention Mechanism

Multi-label text classification is a widely studied subtask in the field of natural language processing. Unlike single-label text classification, multi-label text classification involves assigning multiple labels to a single piece of text. This approach better reflects real-world scenarios. In this study, the authors utilize the news events public dataset from Paddle AI Studio and build upon existing literature to enhance the performance of the existing model for multi-label classification research. Through a comparative analysis with two other models, the authors conclude that incorporating a two-layer convolutional neural network and a self-attention mechanism significantly improves the effectiveness of multi-label classification. In terms of the five evaluation metrics used, the proposed model outperforms the comparison group model in four of them. The accuracy rate of the proposed model reaches 83%, which is three percentage points higher than that of the comparison group model.