电力信息与通信技术2026，Vol.24Issue(5)：40-48,9.DOI:10.16543/j.2095-641x.electric.power.ict.2026.05.05
面向电网客服的大语言模型数据处理和训练
Data Processing and Training of Large Language Model for Power Grid Customer Service
摘要
Abstract
With the rapid development of Internet technology,enterprise digital transformation has become an inevitable trend,intelligent customer service as an important part of the enterprise,its digital upgrade is particularly critical.Traditional intelligent customer service relies on a lot of manpower and material input,and the efficiency is low.In this paper,an intelligent customer service system is designed for power grid field,and a large model data processing and training method is proposed to effectively solve the problem of data acquisition and preprocessing.The system can intelligently reply to common work order problems,quickly and accurately provide solutions or suggestions,significantly improve work order processing efficiency,and reduce manual intervention time and workload.Experimental results show that the system outperforms traditional methods in intent recognition and dialogue success rate,achieving an average improvement of approximately 18.7%.关键词
电网系统/智能客服/大语言模型/数据处理Key words
power grid system/intelligent customer service/large language model/data processing分类
信息技术与安全科学引用本文复制引用
陈磊,王向群,王巍然,何健平..面向电网客服的大语言模型数据处理和训练[J].电力信息与通信技术,2026,24(5):40-48,9.基金项目
国家电网公司双创孵化培育资金项目"基于工单数据的知识服务智能体开发"(SGGR0000ZCJS2400552). （SGGR0000ZCJS2400552）
