摘要
随着对精准高效识别具象技术创新机会的要求日益提高,综合运用词频-逆文档频率(term frequency-inverse document frequency,TF-IDF)算法、LDA(latent dirichlet allocation)主题聚类模型、共现网络分析与多维空间专利地图,进行了技术机会识别与评价。首先,在搜集研究领域的相关专利文献并进行预处理的基础上,运用TF-IDF算法与LDA模型提取了技术主题,确定了创新维度;其次,通过构建共现网络优选了表征创新维度的技术主题词并确定了创新维度,继而与创新法则迭代生成了多维空间专利地图以识别系列潜在技术机会;然后,基于可拓评价法计算了各技术机会的关联度以优选出最佳技术机会;最后,以冷藏集装箱技术机会识别为例,验证了该方法的可行性。结果表明,构建的技术机会识别方法能够为目标企业精准高效、低成本开展技术创新提供可资借鉴的理论参考依据。
Based on the increasingly high demand for accurate and efficient identification of concrete technological innovation opportunities,the TF-IDF algorithm,LDA topic clustering model,co-occurrence network analysis,and multi-dimensional patent map were used comprehensively to identify and evaluate technological opportunities.Firstly,based on the collection and preprocessing of relevant patent literature in the research field,the TF-IDF algorithm and LDA model were used to extract technical topics and determine the innovation dimension.Secondly,by constructing a co-occurrence network,the technical subject words representing the innovation dimension were optimized and the innovation dimension was determined,and then the multi-dimensional spatial patent mapwas iteratively generated with the innovation rule to identify a series of potential technology opportunities.Then,based on the extension evaluation method,the correlation degree of each technical opportunity was calculated to select the best technical opportunity.Finally,the feasibility of this method was verified by taking the opportunity identification of refrigerated container technology as an example.The results indicate that the technology opportunity identification method constructed in this article can provide a theoretical reference for the target enterprise to carry out precise,efficient,and low-cost technological innovation.
作者
王金凤
徐泽
冯奕程
张虎翼
WANG Jinfeng;XU Ze;FENG Yicheng;ZHANG Huyi(China Institute of FTZ Supply Chain,Shanghai Maritime University,Shanghai 201306,China;School of Economics and Management,Shanghai Maritime University,Shanghai 201306,China;Beijing Gaode Yunxin Technology Co.,Ltd.,Beijing,100102;Institute of Logistics Science and Engineering,Shanghai Maritime University,Shanghai 201306,China)
出处
《工业工程与管理》
CSCD
北大核心
2024年第6期12-23,共12页
Industrial Engineering and Management
基金
国家重点研发计划(2022YFF0608700)
国家自然科学基金面上项目(72371155)
上海市科技计划项目(20040501300)。
关键词
技术机会识别
共现网络分析
多维空间专利地图
冷藏集装箱
identification of technical opportunities
co-occurrence network analysis
multi-dimensional spatial patent map
refrigerated container
