【目的】 为解决审稿专家信息更新不及时、编辑凭经验送审等因素导致拒审的问题,提出一种基于向量空间模型(Vector Space Model, VSM)和余弦相似度的稿件精准送审方法。【方法】 首先,结合文献调研和《数据分析与知识发现》送审情况分析拒审的关键原因;其次,在中国知网中获取该刊审稿专家(155人)近5年发表的全部论文(1805篇),并使用词频-逆文档频度(Term Frequency-Inverse Document Frequency,TF-IDF)方法计算关键词权重以构建专家VSM;最后,利用余弦相似度模型为稿件匹配最优的外审专家。【结果】 拒审会导致审稿周期加长,并且外审专家库中活跃的审稿专家减少。实验结果表明,所提方法能够提高稿件送审的准确率。【结论】 所提出的稿件送审方法能够弥补人为匹配的缺陷,降低拒审概率。
[Purposes]This paper aims to propose a method for accurate assignment of manuscript review based on vector space model (VSM) and cosine similarity, and reduce the rate of rejection caused by lack of update of peer reviewers' information and subjectivity of editors. [Methods] We first analyzed the reasons of rejection based on literature research and the submission of Data Analysis and Knowledge Discovery. Then, we collected 1805 papers published by the peer reviewers from CNKI, and constructed VSM for peer reviewers by using term frequency-inverse document frequency(TF-IDF). Finally, we selected the optimal peer reviewers for papers by cosine similarity model. [Finding] The peer review cycles would be delayed as a result of rejection, and active peer reviewers are reducing. The experimental result shows that the proposed method could accurately assign peer reviewers for papers. [Conclusions] The proposed method can make up the defects of artificial matching and reduce the rejection probability.
Vector space model,