【摘要】 文本的关键词识别是文本挖掘中的基本问题之一。在研究现有基于复杂网络的关键词识别方法的基础上,从整个复杂网络拓扑结构特征的信息缺失角度来考察各节点的重要程度。提出强度熵测度来量化评估各节点重要程度,用于解决中文关键词识别问题。实验结果表明,该评估方法简单有效,特别适用于带权复杂网络的节点重要性评估。更多还原

【Abstract】 To identify keywords of the document is one of the fundamental issues for text mining.Focusing on the study of the existing keyword identification approaches based on complex networks,we exploit the importance of nodes from the aspect of information missing in the whole complex network topology.We introduce a novel measurement,called strength entropy,to quantitatively evaluate the importance of nodes and solve Chinese keyword identification problem.Experimental results show that the evaluation method is simple and effective,especially for weighted complex networks.更多还原