基于TextCNN和LightGBM的导游违规行为检测  

Illegal Tour Guide Behavior Detection Based on TextCNN and LightGBM

作　　者：刘昌澍 李响 詹瑾瑜[1,2] 江维 李博智[1] 曹扬 杨瑞 LIU Chang-shu;LI Xiang;ZHAN Jin-yu;JIANG Wei;LI Bo-zhi;CAO Yang;YANG Rui(School of Information and Software Engineering,University of Electronic Science and Technology of China,Chengdu 610054,China;CETC Big Data Research Institute Co.,Ltd.,Guiyang 550022,China;Big Data Application on Improving Government Governance Capabilities National Engineering Laboratory,Guiyang 550022,China)

机构地区：[1]电子科技大学信息与软件工程学院,四川成都610054 [2]中电科大数据研究院有限公司,贵州贵阳550022 [3]提升政府治理能力大数据应用技术国家工程实验室,贵州贵阳550022

出　　处：《计算机技术与发展》2021年第5期143-149,共7页Computer Technology and Development

基　　金：提升政府治理能力大数据应用技术国家工程实验室开放基金项目(W-2019007);四川省科技计划项目(2018CC0136);中科院计算机体系结构国家重点实验室开放课题(CARCH201811);中央高校基本科研业务费(ZYGX2018J077,ZYGX2019J078)。

摘　　要：人工处理旅游评论需要耗费大量人力,如何自动分析旅游评论检测出导游违规行为,为旅游监管提供依据,成为一个迫切需要解决的热点问题。该文提出了一种基于TextCNN和LightGBM的导游违规行为检测方法,首先构建旅游评论的文本卷积神经网络(TextCNN)从海量旅游评论信息中筛选出负面评论;再将这些负面评论送入基于梯度提升决策树(LightGBM)的导游违规行为分类模型,分析得到导游违规行为的具体类型分类及分类概率。使用准确率、召回率、F1值等多个性能指标对提出的模型进行测试与分析,实验数据表明,基于TextCNN和LightGBM的导游违规行为检测方法比一些其他主流方法和模型(SVM、LSTM、XGBoost等)具有更好的准确性和合理性。同时,该方法应用在实际旅游大数据系统中可以得到91.57%的准确率。It takes a lot of manpower to process tourist comments manually.How to process tourist comments automatically and provide the evidence for tourism regulation has become an issue that needs to be addressed urgently.We propose an illegal tour guide behavior detection method based on TextCNN and LightGBM.Firstly,the tour guide negative comment identification model based on text convolutional neural networks(TextCNN)is presented to filter the tourist comments by emotion analysis.Then the filtered negative tourist comments are sent into the illegal tour guide behavior detection model based on light gradient boosting machine(LightGBM)to obtain the corresponding classification.We use precision,recall,F1-score to evaluate the performance of the proposed models.The experiment shows that the proposed illegal tour guide behavior detection method has better accuracy and rationality compared with other models including SVM,LSTM,XGBoost,etc.And the proposed method can achieve 91.57%precision in the practical tourism big data system.

关 键 词：自然语言处理 情感分析 导游违规行为 文本卷积神经网络 梯度提升决策树 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]