基于专利异构网络的技术创新人才发现方法

Technical Innovation Talent Discovery Method Based on Heterogeneous Network

专利数据中包含了丰富的科技信息。通过对专利数据的分析,不仅可以帮助企业快速了解当前领域的发展状况,还可以及时发现领域中的技术创新人才。尽管当前已经存在了一些基于专利指标的技术创新人才评价方法,但这些方法往往孤立地对各个特征进行分析,而忽略了考虑特征之间的关联对技术创新的人才发现的影响。针对该问题,论文提出了一种基于专利异构网络的技术创新人才发现方法,不仅考虑多个特征来评价专利数据中发明人的创新能力,而且通过构建多维特征的异构网络将多维度特征关联起来,从而可以在专利数据集合中准确地发现技术创新人才。实验表明,提出的方法是有效的。

Patent data contains rich scientific and technological information.Through patent analysis,the enterprise can not only quickly understand the development of the given field,but also help to discover technical innovation talents timely.Although there exist some methods of technical innovation talents discovery based on patent indicators,the patent features are always taken as isolated individuals in these methods,and the correlation between patent features are ignored.To solve this problem,this paper proposes a technical innovation talent discovery method based on patent heterogeneous network.It not only consideres many features to evaluate the inventor’s innovation ability in patent data,but also associates multi-dimensional features by constructing patent heterogeneous network,so it can be accurated to discover the technical innovation talents in patent data.Experiments show that the proposed method is effective.