数据挖掘在高职院校就业信息管理中的应用研究

发布时间：2018-08-02 11:33

【摘要】：随着我院毕业生人数的增多,就业工作及就业方向指导越发沉重,研究开发就业信息管理系统,增强我院毕业生的就业水平及就业方向非常有必要。如何利用数据挖掘技术来获取就业指导信息,提高工作效率尤为重要,本文主要从以下三个方面来进行研究：针对就业信息系统中存在的大量数据,通过数据预处理进行分析取样,获取毕业生信息中与就业息息相关的关键信息,例如：综合成绩、英语等级、计算机能力、个人性格取向等一系列数据。针对Apriori算法产生大量候选集和重复扫描数据库的不足,本文利用基于APriori的多维频繁项集改进算法,通过将事务数据库映射为一个布尔矩阵,利用逐层递增的方法来动态分配内存,再利用向量求“与”运算,寻找频繁项集。算法产生的候选集得到大量减少,并将其运用到高校就业管理系统中,较好地缩短扫描时间、节省内存开销,提高工作效率。在关联规则挖掘基础上,利用了C4.5决策树算法构造就业决策树,将已毕业学生按计算机能力、英语等级、综合成绩、政治面貌、性格取向等决策属性进行相应分类,来研究对毕业生的就业类别和方向的影响,从而在提高就业率、就业层次、改进现行培养机制等方面,为管理人员提供决策支持。
[Abstract]:With the increase of the number of graduates in our college, the guidance of employment work and employment direction is becoming more and more heavy. It is necessary to study and develop the employment information management system to enhance the employment level and employment direction of our college graduates. How to use data mining technology to obtain employment guidance information and improve work efficiency is particularly important. This paper mainly studies from the following three aspects: aiming at a large number of data in employment information system, Through data preprocessing, the key information related to employment is obtained, such as comprehensive scores, English grade, computer ability, personality orientation and so on. In view of the shortage of Apriori algorithm producing a large number of candidate sets and repeatedly scanning database, this paper uses the improved algorithm of multidimensional frequent itemsets based on APriori to map the transaction database to a Boolean matrix. The memory is allocated dynamically by layer increments, and the frequent itemsets are found by using vector operation of "and". The candidate set generated by the algorithm has been greatly reduced and applied to the employment management system in colleges and universities, which can shorten the scanning time, save memory cost and improve work efficiency. On the basis of association rule mining, C4.5 decision tree algorithm is used to construct employment decision tree, which classifies graduate students according to computer ability, English grade, comprehensive achievement, political outlook, personality orientation and other decision attributes. In order to improve the employment rate, employment level, improve the current training mechanism and other aspects, provide decision support for managers.
【学位授予单位】：广西大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：G717.38;TP311.13

【参考文献】