基于混合神经网络的电力短文本分类方法研究

Short Text Classification of Electric Power Based on Hybrid Neural Network

随着移动互联网时代的到来,如何保持一个良好的用户群,提高用户的满意度,成为各个领域关注的焦点。客户投诉是直接反馈用户的意见,可以通过多个渠道获得大量的用户投诉数据,从这些数据中获得很多有用信息,而目前的用户投诉往往依靠人工审核,缺乏系统和自动化的投诉分析工具。论文从电力用户投诉短文本入手,对用户投诉的自动分类进行研究,并结合用户的投诉与用户的属性,提取有用的投诉信息。通过这些研究,最终将有助于改善电网综合服务水平。

With the advent of mobile Internet era,how to maintain a good user group and improve user satisfaction have be. come the focus of attention in various fields. Customer complaints are direct feedback to users'opinions. A large number of user com. plaints data can be obtained through multiple channels,and a lot of useful information can be obtained from these data. At present, user complaints often rely on manual review,and lack of systematic and automated complaints analysis tools. Starting from the short text of power users'complaints,this paper studies the automatic classification of users' complaints,and extracts useful complaints in. formation by combining the complaints of users and their attributes. These studies will ultimately help improve the comprehensive service level of the power grid.