An Integrating Emotional Information and Topic Model Method for Text Topic Mining
Chao Chen, Zuqin Chen, Tingkai Hu, Jike Ge, Haoyu Peng, Can Liu · 2021
With the development of natural language processing technology, the topic mining of the text has become very important, because the hidden topic information in the text can significantly enhance the semantic information of the text. We take the neural topic model CombinedTM as the prototype, modify the input of the original model, combine the emotional information of the text with the contextual representation of the text as the new input of the model, add emotional level semantic information to the topic model, and build an improved topic model CombinedETM. Finally, we discuss the performance and practical value of the model with examples. The simulation results show that the topic model fused with emotional information can generate more coherent topics, and can accurately extract the topic information hidden in real events.