Budgetary Shifts vs. Project Volume: A Temporal Text Mining Analysis of South Korean Hydrogen R&D (2002-2024)
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
As global climate change accelerates, hydrogen has emerged as a compelling fossil fuel alternative due to its high energy density and minimal geographical constraints. Many countries are investing in comprehensive hydrogen value chain R&D, and South Korea has intensified government-led initiatives to strengthen its competitive position in the hydrogen economy. This study analyzes domestic hydrogen energy R&D trends using advanced text mining techniques on 10,353 project documents from the National Science & Technology Information Service (NTIS) spanning 2002-2024. The methodology employs Latent Dirichlet Allocation (LDA) topic modeling to extract key thematic areas and semantic network analysis to visualize inter-topic relationships and knowledge structures. The analysis identifies thirteen principal research topics, with hydrogen production technologies and research investment strategies as the most prominent themes. Network analysis reveals six distinct topic clusters, with LDA demonstrating superior topic classification accuracy. This research contributes academically by applying diverse text mining methodologies to hydrogen R&D data from a longitudinal perspective, elucidating the evolutionary trajectory of research priorities. Practically, the study provides insights into temporal shifts in R&D focus areas and funding allocation patterns, offering strategic implications for policy formulation and research planning in the hydrogen energy sector.
Abstract
As global climate change accelerates, hydrogen has emerged as a compelling fossil fuel alternative due to its high energy density and minimal geographical constraints. Many countries are investing in comprehensive hydrogen value chain R&D, and South Korea has intensified government-led initiatives to strengthen its competitive position in the hydrogen economy. This study analyzes domestic hydrogen energy R&D trends using advanced text mining techniques on 10,353 project documents from the National Science & Technology Information Service (NTIS) spanning 2002-2024. The methodology employs Latent Dirichlet Allocation (LDA) topic modeling to extract key thematic areas and semantic network analysis to visualize inter-topic relationships and knowledge structures. The analysis identifies thirteen principal research topics, with hydrogen production technologies and research investment strategies as the most prominent themes. Network analysis reveals six distinct topic clusters, with LDA demonstrating superior topic classification accuracy. This research contributes academically by applying diverse text mining methodologies to hydrogen R&D data from a longitudinal perspective, elucidating the evolutionary trajectory of research priorities. Practically, the study provides insights into temporal shifts in R&D focus areas and funding allocation patterns, offering strategic implications for policy formulation and research planning in the hydrogen energy sector.