基于技术关联关系的企业研发潜在合作伙伴识别方法研究综述

Review of Methods for Enterprises R&D Potential Partners Identification Based on Technology Association Relationship

[目的/意义]技术关联与匹配是企业进行研发合作的内在动因,梳理基于技术关联关系的企业研发潜在合作伙伴识别方法,总结研究不足并提出发展建议,完善面向潜在合作伙伴识别的情报方法体系。[方法/过程]基于136篇重点文献,归纳基于技术承继、共现、结构、应用关系以及复合技术关系的企业研发潜在合作伙伴识别方法,对比分析每种方法的优劣,并提出未来研究方向。[结果/结论]现有研究通过挖掘引用关系、共现关系、文本语义、复杂网络中的有用信息,综合统计和语义特征取得了较好的识别效果。不足在于:数据源和数据范围受限,技术文本语义分析方法存在缺陷,技术关联与合作行为的关系未梳理清晰。未来方向:纳入多类型数据并保障识别范围的完整性;完善技术文本语义分析与计算的理论方法体系;系统梳理技术关联与合作行为的关系,完善合作潜力测度指标。

[Purpose/Significance]Technology association and matching are the fundamental reasons for R&D cooperation.Sort out the identification methods of potential R&D cooperation of enterprises which based on technology association,summarize the research shortcomings and put forward development suggestions,and improve the intelligence method system for potential partner identification.[Method/Process]Based on 136 key literatures,the study summarized the potential cooperative identification methods which based on technology inheritance relationship,co-emergence relationship,structure relationship,application relationship and composite technical relationship,compared and analyzed the advantages and disadvantages of each method,proposed future research directions.[Result/Conclusion]The existing research has achieved good recognition effect by mining useful information in citation relationship,co-occurrence relationship,text semantics,and complex network,combining statistical and semantic features.The disadvantages are:The data source and data scope are limited;The technical text semantic analysis method has defects;The relationship between technical association and cooperative behavior is not clear.Future directions:Incorporate multiple types of data and ensure the integrity of the scope of identification;Improve the theoretical method system of semantic analysis and calculation of technical texts;Systematically sort out the relationship between technical association and cooperative behavior,and improve the cooperation potential measurement index.