A Text Classification Model Based on Gaussian Multi-Head Self Attention Mechanism for Chinese Medical Data
Mo Chen, Chunlong Yao, Xu Li, Lan Shen · 2023
In order to improve the matching level of the Q&A system, text classification methods are used to classify user questions. Therefore, this paper proposes a text classification based on a hybrid model of CNN and Transformer in chinese medical dataset. Unlike the existing traditional CNN-based methods, we use a Gaussian multi-head self attention mechanism to obtain the text’s spatial features. On the chinese medical dataset ,experimental results show that compared with the baseline model the proposed model performs better in accuracy,recall,precision and F1, in which the accuracy achieves 93.20% and the F1 value achieves 94.66%.