融合内容相似度和路径相似性的潜在作者合作关系挖掘

Mining Potential Author Cooperative Relationships Based on the Similarity of Content and Path

[目的/意义]科学论文的作者合作是科研合作的显著表现形式之一,挖掘和识别潜在作者合作关系,有利于推动科研合作、促进科学进步和发展。[方法/过程]在作者关键词耦合分析方法衡量内容相似度的基础上,融入一定权重的链路预测中基于网络结构的路径相似性指标,综合运用作者研究内容相似度和合作网络结构相似性进行潜在作者合作关系挖掘。[结果/结论]以我国生物医学领域近5年的中文期刊论文作者为对象进行实证分析,结果表明该综合挖掘方法效果相对较好且对既有作者合作关系未来发展的可持续性具有预测意义。

[Purpose/significance] Author cooperation based on scientific papers is one of the significant forms of scientific research cooperation.Mining and identifying potential author cooperation is helpful to promote scientific research cooperation and development.[Method/process] On the basis of the author keyword coupling analysis method,this paper integrates the path similarity index in the complex network with certain weight,and constructs a potential cooperative relationship mining method that based on the similarity of the author’s research content and the cooperative network structure.[Result/conclusion] The authors of Chinese journal articles published in recent five years in the field of biomedicine are taken as the research object to make an empirical analysis.The results show that the comprehensive mining method is relatively effective and has predictive significance for the sustainability of the existing cooperative relationship.