Figure 1.
The flow diagram illustrates a multi phase analytical framework for technology analysis using patent data. The process begins with K I P O Patent Database search from 1990 to 2023, followed by Phase 1 Data Acquisition and Refinement including data cleaning, duplicate removal, noise filtering, and text corpus construction. Phase 2 Semantic Text Mining includes natural language processing, tokenization, stop words removal, and feature engineering using T F I D F weighting and keyword frequency analysis, leading to final feature selection of top 20 key technologies. Phase 3 A Trend Dynamics applies stationarity testing and A R I M A model fitting to enable trend classification into Hot, Active, Cold, and Emerging. Phase 3 B Structural Connectivity constructs a co occurrence matrix, measures centrality, and performs community detection. These outputs feed into integrated technology analysis, concluding with Phase 4 Strategic Insights including a technology priority matrix based on compound annual growth rate versus network centrality and development of R and D strategic roadmap and policy implications.Research framework diagram