Customer Satisfaction Research based on Customer Service Dialogue Corpus

Recently, consumers are increasingly inclined to contact customer service for help when they encounter problems and have high demands for remote support. A high-quality customer service connects the company with its customers and establishes a positive image. Applying customer satisfaction metrics to measure the quality and efficiency of customer service is widely used, yet most of the existing customer service evaluation systems rely on manual processes, which are clearly unsustainable and costly. We introduce an ERNIE-based customer satisfaction analysis model that automatically analyses the text of customer service dialogues and scores them from four perspectives (i.e., product, service, process and overall) without human involvement. Furthermore, we construct a corpus containing around 1500 entries of dialogues texts transcribed from customer service consultation and scale it up to 9 times in the training phase. Results show that our model performs better compared to the baseline model and demonstrates a good generalization ability as well.