基于Weka的就业数据分析和模式挖掘——以重庆市信管专业为例

Employment Data Analysis and Pattern Mining Based on Weka——Take Specialty of Information Management and Information System in Chongqing for Example

在Weka平台下使用了时间序列预测算法、朴素贝叶斯聚类和Apriori关联规则算法分析真实就业数据,得到如下结论:该专业各年的男女比例总体较为平衡,很大程度上是因为该专业在重庆市的5所高校中既招收文科生,又招收理科生;女性党员学生且没有工作经验与"考取研究生"这一状态关联度较高;企业是信管专业的最主要的就业领域;获奖次数多的学生选择传统就业领域的概率比较大;该专业中、东部的就业信息向重庆市传递还不够充分。文章还分析了产生这些结论的2个原因:一是图书情报部门已经不能形成一个信息管理与信息系统专业独特的就业类型,但社会对该专业产生了新的需求;二是实际就业环境与专业定位设计存在差距。

We use WEKA as a tool for data analysis and pattern mining; the data come from five universities in Chongqing. Time se- ries forecasting algorithm, decision tree classifier, and naive Bayesian clustering and improved a prior association rules algorithm are used for our research. We analysis some of the visualization of mining results, get some patterns in the data of specialty of informa- tion management and information system in Chongqing. The results of our research include： the proportion of male and female is balanced overall, largely because both liberalists students and science students are recruited; female students with no work experi- ence are high correlation with the state ＂admitted students＂; companies are the leading area of employment of this specialty~ the students get more awards, they choose more traditional job areas; it is not sufficient for job information of eastern China transferring to Chongqing. This paper also analyzes two reasons., first, libraries have been unable to form a unique job type, but new needs are I^enerated at the same time; second, there are ~taD between lob environment and the design of specialty oosition.