［目的/意义］ 针对传统单一地理或属性分析难以解释“地理接近却未合作”等隐性知识关联缺失的困境，系统梳理多维邻近性从单一维度到静态五维框架，并向动态化、情境化演进的逻辑，厘清其现有的测度方法体系。在此基础上，进一步推动该理论从概念阐释工具向融合多源数据与人工智能的智能决策支持系统演进，以期为优化科技创新布局及提升创新体系韧性提供坚实的理论依据与实践路径。［方法/过程］ 基于CNKI与Web of Science数据库相关文献，综合运用文献计量、内容分析法与关键词共现方法，追溯邻近性理论从地理决定论向多维动态框架的演进过程，归纳单一维度测度、多维整合测度与智能化测度方法，并结合VOSviewer图谱结果分析其在情报学中的主要研究热点与应用主题。［结果/结论］ 研究表明，多维邻近性测度已呈现出从静态单维向动态多维体系演进的总体趋势，复杂情境下的交互机制解析成为方法创新的核心。同时，数据驱动与智能技术正深刻重塑该领域的分析路径。尽管现有研究在指标体系的完备性与动态适应性方面仍面临挑战，但构建融合多源信息的智能化分析框架将是突破当前瓶颈、提升创新关系识别与决策支持能力的关键方向。

[Purpose/Significance] Addressing the dilemma where traditional single-dimensional geographic or attribute analyses fail to explain the absence of implicit knowledge connections (e.g., "geographic proximity without collaboration"), this paper aims to systematically review the logical evolution of multidimensional proximity theory—from a single dimension to a static five-dimensional framework, and further towards dynamic and contextualized paradigms. It also clarifies the existing system of measurement methods. Building on this, the study seeks to promote the theory's evolution from a conceptual explanatory tool into an intelligent decision-support system integrating multi-source data and artificial intelligence, thereby providing a solid theoretical basis and practical pathways for optimizing the layout of technological innovation and enhancing the resilience of innovation systems. [Method/Process] Based on relevant literature retrieved from CNKI and Web of Science databases, this study integrated bibliometric analysis, content analysis, and keyword co-occurrence analysis. It traced the evolution of proximity theory from geographical determinism to a multidimensional dynamic framework, summarized measurement approaches including single-dimensional measurement, multidimensional integrated measurement, and intelligent measurement methods, and, in combination with VOSviewer mapping results, analyzed the main research hotspots and application themes of multidimensional proximity theory in the field of information science. [Result/Conclusion] The research indicates that the measurement of multidimensional proximity shows a general trend of evolving from static single-dimensional to dynamic multi-dimensional systems, with the analysis of interaction mechanisms in complex contexts becoming the core of methodological innovation. Meanwhile, data-driven approaches and intelligent technologies are profoundly reshaping analytical pathways in this field. Although current research still faces challenges regarding the completeness of indicator systems and dynamic adaptability, constructing an intelligent analysis framework that integrates multi-source information is the key direction for breaking through current bottlenecks and enhancing capabilities in identifying innovation relationships and supporting decision-making.