Research on Interactive Text Topic Mining Based on LDA Model -- Taking Customer Service Chat Records as an Example

The paper is to dig out the topics contained in customer service chat records, which can provide guidance for the design and optimization of customer service automatic question answering system. In order to mine the interactive short texts of customer service chats, preprocessing and feature selection are carried out in the paper. Finally, the topics are obtained by LDA model. The results of the model show that topic customers concerned about are in line with the actual situation. Different topics are independent of each other, at the same time, there is a strong correlation between topics and sub-topics. In the process of text mining, there are some cross-links in different word filtering methods. The LDA modeling results provide guidance for the design and optimization of automated QA systems for e-commerce websites. In addition, the paper puts forward a new avenue for web users to understand the topics of concern.