基于Hadoop的高校学生行为预警决策系统研究

通过智慧校园各种智能终端、可感知设备,获取学生校园动态行为轨迹数据,构建多维数据存储中心。利用Hadoop框架分布式文件系统HDFS和MapReduce,将多维数据进行关联、分类、降维、聚类分析与可视化呈现。使学生特征标签化,生成基于特征矩阵的学生画像,从而分离出偏离中心点的学生异常,建立动态的预警决策机制,从而进行前置预警和智慧决策;使高校管理机构及教师主动掌握学生的生活情况、学习状态及行为规律,从而对不良思想行为做到事先警示教育、事后跟踪管理,实现以学生为视角的智能管控及智慧管理。基于Hadoop的预警决策系统,开创了智慧校园教育管理决策科学化、管理智能化、监督过程化的新模式,具有较高的经济效益与推广应用价值。

HADOOP-BASED COLLEGE STUDENT BEHAVIOR WARNING DECISION SYSTEM

By means of various intelligent terminals and perceptible devices in the smart campus,the dynamic behavior trajectory data of the student campus are obtained,and the multi-dimensional data storage center is constructed.The Hadoop framework distributed file system HDFS and MapReduce are used to correlate,classify,reduce dimension,analyze cluster and visualize the multi-dimensional data.It can label the characteristics of students,generate the students’portraits based on the characteristic matrix,separate the students’anomalies that deviate from the center point,establish a dynamic early-warning decision-making mechanism,thus carrying out early-warning and wise decision;it can make the management institutions and teachers in colleges and universities take the initiative to grasp the life situation of the students,learning state and behavior rule,and bad thoughts and actions do advance warning education,after the event tracking management,implementing the perspective of students of the intelligent control and intelligent management.The early warning decision system based on Hadoop creates a new model of scientific decision making,intelligent management and process supervision for intelligent campus education management,which has high economic benefits and application value.