电力文本数据挖掘现状及挑战  

Current Status and Challenges of Power Text Data Mining

作　　者：王慧芳[1] 曹靖[1] 罗麟 WANG Huifang;CAO Jing;LUO Lin(College of Electrical Engineering, Zhejiang University, Hangzhou 310027, China;State Grid Zhoushan Power Supply Company, Zhoushan Zhejiang 316021, China)

机构地区：[1]浙江大学电气工程学院,杭州310027 [2]国网浙江省电力有限公司舟山供电公司,浙江舟山316021

出　　处：《浙江电力》2019年第3期1-7,共7页Zhejiang Electric Power

基　　金：国网浙江省电力有限公司群创项目(5211ZS180011)

摘　　要：文本数据是电力大数据的重要组成部分,对其进行有效挖掘是智能电网深入、全面发展的需要。在目前已有研究成果的基础上,对电力领域文本数据挖掘的现状及挑战进行了深入剖析。首先分析了文本挖掘技术的发展过程及存在难题;接着重点分析了电力文本数据挖掘的关键技术及其研究现状,包括文本预处理技术、文本表示方法以及数据挖掘方法;然后以电力设备缺陷文本为对象,介绍了文本挖掘技术在电力领域的应用,包括缺陷文本质量的提升与保证、缺陷文本严重程度自动分类、缺陷发生部件及程度的自动提取、缺陷文本检索、基于缺陷文本的电力设备健康状态评价等,可为其他类型电力文本的挖掘提供参考;最后,探讨了电力文本挖掘面临的挑战以及未来发展方向。Text data is an important part of power big data. It is the need for deep and comprehensive development of smart grid to explore text data effectively. Based on the existing research results, the current status and challenges of text data mining in the power field are deeply analyzed. Firstly, the development process and difficulties of text mining technology are analyzed. Then the key technologies and research status in the process of power text data mining are focused on, including text preprocessing technique, text representation method, and data mining method. Based on the current power equipment defect texts, the paper introduces the application of text mining technology in the power field, including improvement and guarantee of defect text quality, automatic classification of defect text severity, automatic extraction of defect occurrence components and degrees, defect text retrieval, power device health status evaluation based on defect text and so forth. These applications provide reference for mining other types of power texts. Finally, the challenges and future development directions of power text mining are discussed.

关 键 词：电力大数据 文本数据 数据挖掘 自然语言处理 电力设备 缺陷文本 

分 类 号：TM73[电气工程—电力系统及自动化]