基于客服聊天记录的问答语料标注系统设计

摘要

为了解决自动问答系统中对买家的提出问题自动分类，在分析买家聊天记录基础上，运用改进的朴素贝叶斯分类算法，构建了结构化的语料库。通过设计的语料标注系统可完成买家聊天记录的自动分类，测试结果表明，改进后的文本分类算法提高了对问题自动回答的准确度。%In order to automatically categorizes for the buyers questions in the auto Q&A system, the paper analyzes the buyers chat and designs the structured corpus based on the improved Naive Bayesian classification algorithm. The results indicates that the corpus tagging system can automatically categorizes for chat, and enhanced the auto answer accuracy based on the improved text classification algorithm.