Bidirectional LSTM with Hierarchical Attention for Text Classification
Jianping Li, Yimou Xu, Huaye Shi · 2019
The text classification is a basic task of natural language processing (nlp), which aims to get the corresponding category labels for texts with multiple categories. Nowadays, neural network models have been widely used in the nlp field and has achieved remarkable results in text classification. However, due to the high dimensionality of text data and the complex semantics of natural language, there are still many areas for improvement in the network structure of text classification. In order to cope with the above problems, this paper proposes a new network structure, which includes bidirectional long short-term memory (LSTM) combined with hierarchical attention mechanism. In this network structure, the data is sent to bidirectional LSTM after one-hot encoding, and the output is subjected to hierarchical attention. Finally, the softmax classifier is used to classify the processed context information. The experimental results show that the structure has high accuracy in text classification.