LightXML: Transformer with Dynamic Negative Sampling for High-Performance Extreme Multi-label Text Classification
This work provides a more efficient and accurate solution for extreme multi-label text classification, which is beneficial for applications dealing with very large label sets.
This paper addresses the challenge of Extreme Multi-label text Classification (XMC) by proposing LightXML, a new method that uses end-to-end training and dynamic negative label sampling. LightXML achieves state-of-the-art performance on five XMC datasets while significantly reducing model size and computational complexity, for example, reducing model size by up to 72% on the Amazon dataset with 670K labels.
Extreme Multi-label text Classification (XMC) is a task of finding the most relevant labels from a large label set. Nowadays deep learning-based methods have shown significant success in XMC. However, the existing methods (e.g., AttentionXML and X-Transformer etc) still suffer from 1) combining several models to train and predict for one dataset, and 2) sampling negative labels statically during the process of training label ranking model, which reduces both the efficiency and accuracy of the model. To address the above problems, we proposed LightXML, which adopts end-to-end training and dynamic negative labels sampling. In LightXML, we use generative cooperative networks to recall and rank labels, in which label recalling part generates negative and positive labels, and label ranking part distinguishes positive labels from these labels. Through these networks, negative labels are sampled dynamically during label ranking part training by feeding with the same text representation. Extensive experiments show that LightXML outperforms state-of-the-art methods in five extreme multi-label datasets with much smaller model size and lower computational complexity. In particular, on the Amazon dataset with 670K labels, LightXML can reduce the model size up to 72% compared to AttentionXML.