基于文本相似度的简历匹配推荐算法研究

Research on Resume Matching Recommendation Algorithm Based on Text Similarity

为了能够更好的帮助求职者选择合适的岗位信息,提出了基于文本相似度的简历匹配推荐算法。通过对简历和岗位信息进行特征提取及特征处理,将数据文本划分为两类。在结构化文本中,利用偏好权重因子α消除求职者与企业之间由于不同因素带来的相似度计算偏差。对于非结构化文本,利用机器学习模型doc2vec训练,并计算两者之间的相似度,提出修正参数χ来解决doc2vec缺乏考虑段落长度差异性的问题。实验结果表明,上述方法具有更高的准确率、召回率以及F1值。

In order to help job seekers to right select appropriate jobs, this paper proposes a resume recommendation algorithm based on text content similarity. After feature extraction and feature processing, Data texts are divided into two parts. In the structured text, a preference weight factor α is proposed to eliminate the deviation caused by different preferences between users and enterprises. In the unstructured text, a modified parameterχis proposed to solve the problem that calculating similarity using the machine learning model doc2evc caused by the lack of consideration the of length deviation between two paragraphs. The experimental results show that this paper’s algorithm has higher accuracy, recall and F1 value.