Semantic-Unit-Based Dilated Convolution for Multi-Label Text Classification
This work addresses multi-label text classification, an incremental improvement for tasks like document categorization.
The authors tackled multi-label text classification by proposing a model that uses semantic-unit-based dilated convolution and a hybrid attention mechanism, achieving significant advantages over baseline models on RCV1-V2 and Ren-CECps datasets and showing robustness for low-frequency labels.
We propose a novel model for multi-label text classification, which is based on sequence-to-sequence learning. The model generates higher-level semantic unit representations with multi-level dilated convolution as well as a corresponding hybrid attention mechanism that extracts both the information at the word-level and the level of the semantic unit. Our designed dilated convolution effectively reduces dimension and supports an exponential expansion of receptive fields without loss of local information, and the attention-over-attention mechanism is able to capture more summary relevant information from the source context. Results of our experiments show that the proposed model has significant advantages over the baseline models on the dataset RCV1-V2 and Ren-CECps, and our analysis demonstrates that our model is competitive to the deterministic hierarchical models and it is more robust to classifying low-frequency labels.