基于学生多源数据的跨域关联和数据分析

摘要

There is an increasing number of big data related to college students in terms of types and quantitie. However, most of the multi-source data derives from various systems of different departments which has wide inter-structural differences and lacks effective data intermediation and data sharing. To solve the problem, this paper collects multi-sourced student data, including the course score from the Teaching Department, basic information from the student department and the student ID card records from the logistics department. After a first-step data processing and data structuring, the article carried out a cross-domain correlation and further data analysis. Through the empirical analysis of small sample groups, results suggested that students with learning difficulties have more influence on the group than the students with better grades; students with regular school canteen dinner performs better academicly; and young-aged students generally achieve better grades within the same group.%高校学生的相关大数据的类型和数量日益增长,但这些多源数据来源于不同部门的各类系统,结构异化,缺乏有效的数据融通与数据共享.针对这一问题,文章采集了教学部门的课程成绩、学生部门的学生基本信息和后勤部门的校园卡刷卡记录,对这些多源数据进行预处理、结构化的操作后,对其进行跨域关联和数据分析.文章通过小规模样本群体的实证分析,得到学困生对群体的影响比学优生更大、经常在学校食堂晚餐的同学课程成绩更佳、在同一群体中年龄小的成绩表现更优等分析结果.