首页 > 硕士 > 工学 > 正文

高校毕业生就业推荐系统的设计与开发

The Design and Implement of Graduate Occupation Recommending System

作者: 专业:计算机应用技术 导师:林鸿飞 年度:2010  院校: 大连理工大学

Keywords

        近年,随着高校毕业生数量持续增长和全球金融海啸给我国经济带来的不利影响,高校毕业生的就业形势日趋严峻。而当前我国各所高校的毕业生就业工作尚不足以为每一名毕业生提供准确有效的就业指导和就业推荐,各高校的就业网更多是仅提供招聘信息发布功能,并不具备信息推荐功能。“高校毕业生就业推荐系统”的设计开发则刚好填补了这份空白。通过“高校毕业生就业推荐系统”,毕业生可以根据自己的个体情况,得到一份科学、可靠的就业推荐,并以此作为择业依据。针对现有网络求职平台在就业推荐过程中存在的缺陷,同时结合高校毕业生求职和企业校园招聘的特点,我们设计了“高校毕业生就业推荐系统”。在系统的设计过程中,我们通过比较应届和往届毕业生基本特征,分别采用基于经验公式和基于SimRank算法两种办法来获得两名学生之间的相似度。随后,根据学生之间的相似度,通过K-Means算法对学生进行聚类分析,并通过进一步分析得到应届毕业生与企业间的相似度。最后,本文将学生与企业的相似度同基于PageRank算法获得的各个企业的“求职指数”结合,从而获得企业的推荐排序权值,并根据这个权值将排序靠前的企业推荐给对应的应届毕业生。尽管本文采用了两种不同的学生相似度计算方法,但通过本文第五章的测试对比实验,在最终系统中,我们选择基于经验公式计算学生间相似度的方法来完成学生间相似度计算。根据测试实验的结果,我们认为:本系统不仅功能上符合设计初衷,能够有效的为毕业生提供就业推荐服务,推荐结果科学合理;同时还能够帮助缺乏求职目标的学生制定求职目标,提升学生的求职成功率,在一定程度上降低学生求职成本。对比之前就业网单纯的信息发布功能,本系统提供的就业推荐功能具有较高的实际应用价值。
    In recent years, because of the continuously growing number of the college graduates and the adverse impact on China’s economy caused by the global financial tsunami,the college graduates faced an increasingly tough job market. However, at present, the graduates-employment service of all the universities in China cannot fully provide each graduate with proper and effective employment guidance and job recommending. Besides, what universities’employment networks have is only recruitment information releasing function, but not information recommending function. The design and development of "the employment recommending system for the college graduates" can just fill the gap. By using this system, the college graduates can get a scientific and reliable employment recommendation according to their individual circumstances. With the recommendation, the graduates can make a wiser choice for their career.As the existing network platform in the employment recommending process has flaws, we have designed "the employment recommending system for the college graduates", which has taken the features of graduates’seeking jobs as well as campus recruitment into consideration. In the system design process, we compare the basic features of fresh and previous graduates and get the similarity of these two groups of graduates by using the empirical formula and Simrank algorithm respectively. Then, based on the result, we obtain the similarity of the fresh graduates and enterprises in further cluster analysis. Finally, we get all enterprises’ recommendation-ranking weight by combining the similarity (of the fresh graduates and enterprises) with the enterprise’s "Job Index" which is obtained by PageRank algorithm. Thus, several top enterprises in our rankings will be recommended to the graduates.Though this article has applied two different algorithms for calculating the similarity between graduates, we choose the empirical formula in the final system according to the result from the comparative testing experiment in Chapter V of this article. We can conclude from the testing experiment result that the final system not only meet the original intention in function, which can effectively offer scientific and reasonable employment recommendation to the graduates, but also help those who lack job objectives to set one and enhance the success rate in seeking a job, which to some extent means reducing the cost in job seeking. Compared with the simple information releasing function of the present employment network, the employment recommendation function in our system has a higher practical value.
        

高校毕业生就业推荐系统的设计与开发

摘要4-5
Abstract5
1 绪论9-12
    1.1 大学生职业推荐系统的开发背景及意义9-10
    1.2 目前国内相关职业推荐平台的调研分析10-11
    1.3 本文主要工作11
    1.4 本文的篇章结构11-12
2 相关技术背景12-24
    2.1 几种推荐方法的介绍12-13
    2.2 SimRank算法介绍13-15
    2.3 聚类分析介绍15-17
    2.4 K-Means算法描述17-18
    2.5 PageRank算法概述18-24
        2.5.1 PageRank算法简介18-19
        2.5.2 PageRank算法思想19
        2.5.3 PageRank算法公式19-21
        2.5.4 PageRank算法计算过程21-24
3 系统概述24-29
    3.1 系统整体设计思路24-25
    3.2 系统模块介绍25-27
        3.2.1 数据预处理模块25
        3.2.2 企业信息抽取模块25-26
        3.2.3 学生间相似度计算模块26
        3.2.4 往届学生聚类分析模块26
        3.2.5 学生与企业间相似度计算模块26-27
        3.2.6 企业"求职指数"计算模块27
        3.2.7 最终权值计算模块27
    3.3 系统环境27-29
        3.3.1 系统硬件环境要求27-28
        3.3.2 系统的软件环境28-29
4 关键模块技术29-42
    4.1 基于经验公式的学生相似度计算29-30
    4.2 基于SimRank算法的学生相似度计算30-35
    4.3 基于K-Means的学生聚类分析35-36
        4.3.1 对往届毕业生进行聚类分析的目的35
        4.3.2 本文对往届学生进行K-Means聚类分析的具体方法35-36
        4.3.3 本文对往届学生进行K-Means聚类分析的结果分析36
    4.4 应届学生与企业之间的相似度计算36-37
    4.5 基于PageRank算法的企业求职指数计算37-41
        4.5.1 企业"求职指数"的相关描述37-38
        4.5.2 基于PageRank算法的企业"求职指数"(PR)计算38-41
    4.6 最终排序权值W计算41-42
5 测试及运行分析42-49
    5.1 系统测试数据的选取42
    5.2 测试衡量标准42-43
    5.3 系统测试环节的设计43
    5.4 系统测试结果分析43-47
        5.4.1 基于经验公式计算学生间相似度方法的系统测试结果分析43-45
        5.4.2 基于SimRank算法计算学生间相似度方法的系统测试结果分析45-46
        5.4.3 对比实验结论46-47
    5.5 系统运行实例及评价47-49
结论49-50
参考文献50-52
攻读硕士学位期间发表学术论文情况52-53
致谢53-55
        下载全文需55


本文地址:

上一篇:基于社会化标注的查询扩展技术研究
下一篇:宽范围输入逆变器的研制

分享到: 分享高校毕业生就业推荐系统的设计与开发到腾讯微博           收藏
评论排行
频道总排行
频道本月排行
(); var h = cur.height(); $(this).find('.img-wrap img').LoadImage(true, w, h,'http://www.fabiao.net/statics/images/msg_img/loading.gif'); }); }) var _bdhmProtocol = (("https:" == document.location.protocol) ? " https://" : " http://"); document.write(unescape("%3Cscript src='" + _bdhmProtocol + "hm.baidu.com/h.js%3F4fe77876a18891357653cdbbd216a6a3' type='text/javascript'%3E%3C/script%3E"));
-高校毕业生就业推荐系统的设计与开发-ѯ