基于加权网随机区块模型的学术热点提取算法

学术热点在把握科学前沿、掌握学术动向、制订科研规划、学术作品评审等领域中有广泛的应用。针对学术热点发现中对文献市场选择性影响体现不足和热点内容结构表现单一的问题,本文提出了以学者选读文献为基础的学术热点提取模型和算法,设计了基于加权网模块社群挖掘算法的随机区组模型两阶段算法,用于发现带结构的学术热点,模拟和实证研究均表明算法在学术热点提取中取得良好效果。

Academic Hotpots Extraction Algorithm Based on Weighted Network Random Block Model

Academic hotpots are widely applied in many academic management tasks including mastering academic trends, funding policies and national research programs. Focus on the lack of literature market effect in hotpots discovery algorithms and the problems of simple contents expression, this paper proposes hotpots extraction algorithm based on papers selected by readers. We design randomized block model based on WFN used for stable result extraction which is called a two-stage approach. The simulation and empirical studies both show that the algorithm can achieve better results in the hotpots extraction.