基于改进TF-IDF算法的毕业生就业推荐算法研究

Research on Graduate Employment Recommendation Algorithm Based on Improved TF-IDF Algorithm

针对传统就业推荐算法不能够对每一个毕业生进行精准的推荐的局限性,论文提出一种结合TF-IDF算法和K-means++算法的双向推荐系统,一方面对毕业生信息使用K-means++算法进行聚类,对新用户根据其初始信息与行为信息进行用户画像建模,并计算与往届毕业生的相似度;另一方面使用TF-IDF算法对各个招聘网站所发布的招聘信息中的关键词进行统计转换词频等操作。实验结果表明,该双向就业推荐系统比起之前单向就业推荐提高了毕业生就业推荐的满意度,提升推荐效率。

In view of the limitation that traditional employment recommendation algorithms cannot accurately recommend every graduate,this paper proposes a two-way recommendation system combining TF-IDF algorithm and K-means++algorithm.On the one hand,K-means is used for graduate information.The means++algorithm performs clustering,models new users based on their initial information and behavioral information,and calculates the similarity with previous graduates,on the other hand,the TF-IDF algorithm is used in the recruitment information released by various recruitment websites perform operations such as statistical conversion of the keywords of the keywords.The experimental results show that the two-way employment recommendation system improves the satisfaction of graduates'employment recommendation and improves the recommendation efficiency compared with the previous one-way employment recommendation.