摘要
文中提出了一个面向电信客户投诉和建议的自动快速分析模型,通过网络爬虫采集互联网上客户的投诉和建议,通过客服录入获取客户的电话投诉和建议信息,然后对这些投诉和建议文本进行预处理,包括中文分词、消去停用词、构建向量空间模型等。最后使用结合了人工信息和统计信息的自动识别算法对客户的投诉和建议进行分析、识别,把它们转交给相应的部门处理。
This paper proposes an intelligent analysis model for telecommunication customer complaints. First, it collects customer complaints from Internet by web crawler and enterprise database typed in by customer service staff. Then, these Chinese texts are processed with Chinese word segment, stop words elimination and space vector model construction. Finally, customer complaints are classified by an intelligent analysis algorithm combined with an artificial complaint dictionary and statistical information. This model can help for improving customer satisfaction and loyalty, reducing loss of customer and operating cost.
出处
《现代电信科技》
2013年第5期4-7,共4页
Modern Science & Technology of Telecommunications
关键词
投诉分析
文本处理
机器学习
complaint analysis, text preprocessing, machine learning
