大数据背景下的学生孤独预警模型——以华中师范大学为例

Early Warning Model of Students’ Loneliness under the Background of Big Data——Taking Central China Normal University for Example

大数据时代背景下,关注大学生心理健康,要借用技术手段科学、客观推进大学生心理健康教育工作。目前研究学生心理健康大多采用问卷调查的形式,所得结果取决于被调查者的填写情况,不一定真实反映其内心的想法,并且调查个案有限,不能很好地反映总体情况。本文以华中师范大学为例,通过一卡通消费明细、图书门禁明细得出学生间的共现频率,从而得出学生的朋友关系表,得到疑似孤独者名单,结合学生的就业情况,得出朋友关系较少的学生未就业率高于朋友关系多的结论;再运用DecisionTreeClassifier模型,挖掘出各个指标对就业成功的影响力,并基于训练好的模型预测哪些学生有就业失败的可能,可作为重点关注对象。分析结果与日常生活反馈较一致,对于应用大数据在高校学生管理工作有一定的借鉴作用。

Under the background of the era of big data,paying attention to the mental health of college students,it is necessary to use scientific means to scientifically and objectively promote the mental health education of college students.At present,the research of students’mental health mostly adopts the form of questionnaire survey,which depends on the filling of the respondents,not necessarily reflect their inner thoughts,and the investigation cases are limited,which can not reflect the general situation well.This paper takes Huazhong Normal University as an example,through the details of the consumption of smartcard and the access details of the library to get the list of students friendship,and get the list of suspected lonely students.Combined with the employment situation of students,it is concluded that the unemployed rate with fewer friends is higher than that of friends;using the DecisionTreeClassifier decision tree model to discover the influence of various indicators of employment success,and predicting which students fail in employment based on the trained model can be the focus of attention.The analysis results are consistent with the daily life feedback,and it has certain reference for the application of big data in the management of college students.