基于大数据的大学生兴趣爱好特征聚类研究  

Research on the cluster analysis of college students’ interests and hobbies based on big data

作　　者：贺奇 董延华[1] 宋嘉怡 王瑜 HE Qi;DONG Yanhua;SONG Jiayi;WANG Yu(College of Computer,Jilin Normal University,Siping Jilin 136000,China)

机构地区：[1]吉林师范大学计算机学院,吉林四平136000

出　　处：《智能计算机与应用》2020年第10期44-46,53,共4页Intelligent Computer and Applications

基　　金：教育部科技发展中心项目(2018A01025);吉林省教育厅科技发展项目(JJKH20191001KJ)。

摘　　要：大学生在社交平台上直接或间接的数据交互,体现了其自身的行为特征,具有重要的分析价值。利用大数据技术对大学生社交平台的交互数据收集、分析及处理,揭示数据背后的大学生兴趣爱好规律,精确对大学生行为进行描述显得尤为重要。本文借助大数据的优势,结合大学生社交平台数据特点,通过虚拟编码、均值填补法和Z-score标准化方法等数据预处理技术,利用k-means聚类算法聚类分析,对具有相似特质和兴趣爱好的大学生进行了分类,划分出了五大类别群体,并且对不同类别的大学生进行了特征分析,为舆情分析、研究大学生群体的兴趣关注点提供数据和理论支持。The direct or indirect interactive data of college students on the social platform reflects their own behavior characteristics and has important analytical value. The use of big data technology to collect,analyze and process the interactive data of college students’ social platform reveals the law of college students’ interests and hobbies behind the data,and it is particularly important to accurately describe college students’ behaviors. In this paper,with the help of the advantage of big data,combining with the characteristics of college students’ social network data,through the virtual coding,the mean filling method and standardization of Z-score data pretreatment technology,such as using k-means clustering algorithm,with similar qualities and interests of college students has carried on the classification,analysis the characteristic of different types of college students,for college students’ group classification,public opinion analysis,corporate advertising,production,marketing to provide data and theoretical support.

关 键 词：大数据 K-MEANS Z-SCORE 聚类分析 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]