信息资源管理学报2026，Vol.16Issue(2)：82-97,16.DOI:10.13365/j.jirm.2026.02.082
语义与演化视角下的产业技术壁垒识别研究
Research on Industrial Technology Barrier Identification from a Semantic and Evolutionary Perspective:The Case of Lithography
摘要
Abstract
Dynamic identification of industrial technology barriers is a strategic cornerstone for breaking the core technology blockades and building secure and controllable industrial chains.Relying on patent texts and control lists,this study proposes an industry technology barrier identification method based on semantic and evolutionary perspec-tives.First,the LDA and Word2Vec models are combined to extract core technical topics,enhancing the accuracy of topic recognition.Then,a three-level screening mechanism based on technology topic identification,technology com-petition mapping,technology control association is constructed to pinpoint key industrial technology barriers.Finally,the Dynamic Topic Model(DTM)is applied to depict the temporal evolution of barrier topics and predict future trends.Taking the field of lithography as a case study,empirical results validate the accuracy and robustness of the proposed method:the results are highly aligned with national strategic priorities,and its evolution trajectories strongly corre-spond with multi-dimensional factors such as industrial milestones,regional policy incentives,and has significant ref-erence value for policy planning and research layout.This study breaks through the limitations of traditional static anal-ysis by constructing a full-cycle research paradigm of"feature identification-mechanism analysis-trend prediction",providing data-driven decision support for overcoming technological bottlenecks in the lithography industry.The pro-posed methodological framework can be extended to strategic domains such as semiconductor equipment and biophar-maceuticals for technical competition and defense studies.关键词
产业技术壁垒/动态主题模型(DTM)/光刻机/语义与演化/演化趋势预测Key words
Industrial technology barriers/Dynamic Topic Model(DTM)/Lithography/Semantics and evolution/Evolution trend prediction分类
社会科学引用本文复制引用
冉从敬,程凡,李旺,蒋云龙..语义与演化视角下的产业技术壁垒识别研究[J].信息资源管理学报,2026,16(2):82-97,16.基金项目
本文系国家自然科学基金面上项目"基于图卷积神经网络的新兴技术领域高质量专利识别及其演化研究"(72274084) （72274084）
山东省自然科学基金青年项目"基于专利计量与机器学习的校企技术合作供需智能匹配方法研究"(ZR2023QG105)的研究成果之一.(This work is supported by the National Natural Science Program of China,"Identification of High Quality Patents in Emerging Technologies Based on Graph Convolutional Neural Networks and Its Evolutionary Study"(72274084)and the Shandong Province Natural Science Foundation Youth Project"Research on Intelligent Matching Method of Supply and Demand for University-Enterprise Technology Cooperation Based on Patent Measurement and Machine Learning"(ZR2023QG105).) （ZR2023QG105）
