摘要
分析了当前国内就业形势,针对当前"招工难"与"就业难"问题并存的现状,提出了一个基于向量相似度的招聘就业双向推荐模型。模型首先按条件对候选推荐信息进行筛选;然后将招聘和求职信息转化为向量,为不同分量建立相应的量化规则并进行量化,使之可计算;最后采用夹角余弦公式计算向量间的相似度,并以此作为双向推荐的标准。模型在测试数据集和实际数据集上均取得了较好的运行效率,准确率高,达到了最优化推荐,一定程度上缓解了江门市目前招聘就业困难的压力,取得了良好的社会效益。
Analysis of the current situation of domestic employment, for the current ＂recruitment difficulties＂ and ＂Employment＂ issues exist status, this paper proposed a bidirectional recommended model of recruitment and employment base on the vector similarity. Firstly, this model screened candidate recommendation information by the conditions, and then converted the recruitment and job information into a vector, for the different components of the vector to create different quantization rules, so made the vector computable; lastly, calculated similarity between vectors by using the cosine formulas, and as the standard of bidirectional recommended. This model has achieved good operating efficiency and high accuracy in both test data sets and real data sets, achieve the optimal recommended, some extent eased the currently pressure of recruitment employment in Jiangmen city and achieved good social benefits.
出处
《中国科技信息》
2013年第21期174-179,共6页
China Science and Technology Information
基金
国家自然科学基金(11271040)
广东省自然科学基金(S2011010003681
S2012010010610)
中国科学院计算技术研究所智能信息处理重点实验室开放课题基金(No.LIP2010-5)
广东省科技计划项目(2010B010600039)
江门市科技计划项目(20120030096094
20120030093981
20130100230255)
五邑大学博士起动基金资助项目(30813008)
关键词
向量
相似度计算
招聘就业
双向推荐
vector
similarity computing
recruitment employment
bidirectional recommended
