Text Mining Strategy of Power Customer Service Work Order Based on Natural Language Processing Technology

Abstract

In order to improve the level of power customer service, we need to start with natural language processing technology, and conduct in-depth text mining of power customer complaint work order. Combined with the format and characteristics of power customer service work order, this paper proposed a text mining strategy of power customer service work order based on natural language processing technology which includes the following processes: work order data cleaning, text segmentation, information characterization, model training and model evaluation is proposed. The TF-IDF algorithm and the model optimization method based on accuracy calculation are applied in this strategy, which greatly improves the level of power customer service.