基于文本分类的高校行政服务质量提升

摘要

基于同济大学“同心云”平台,运用朴素贝叶斯、支持向量机(SVM)、神经网络等机器学习算法,把平台上的微博内容按管理职能部门自动分类.对比几种算法的分类效果,研究停用词、神经网络隐层节点数对分类效果的影响.通过文本分类方法,将广大师生提出的建议和意见及时反馈到相关部门,从而提高高校行政服务质量.%Based on Tongji University's Tongxinyun platform,naive Bayesian,support vector machine (SVM) and neural network are used to classify the microblogs by departments.Then,the effects of the algorithms on the classification are compared,and the effects of disable word and node number of hide layer on classification results are also studied.With the method of text classification,the recommendations and comments of teachers and students are sent to the relevant departments timely,which improves the quality of administrative services in colleges and universities.