Iskanje po katalogu digitalne knjižnice

Opis: : The expectations for the (re)use of open government data (OGD) are high. However, measuring their impact remains challenging, as their effects are not solely economic but also long-term and spread across multiple domains. To accurately assess these impacts, we must first understand where they occur. This research presents a structured approach to developing a taxonomy for open government data (OGD) impact areas using machine learning-driven topic modeling and iterative taxonomy refinement. By analyzing a dataset of 697 OGD use cases, we employed various machine learning techniques—including Latent Dirichlet Allocation (LDA), Non-Negative Matrix Factorization (NMF), and Hierarchical Dirichlet Process (HDP)—to extract thematic categories and construct a structured taxonomy. The final taxonomy comprises seven high-level dimensions: Society, Health, Infrastructure, Education, Innovation, Governance, and Environment, each with specific subdomains and characteristics. Our findings reveal that OGD’s impact extends beyond governance and transparency, influencing education, sustainability, and public services. Our approach provides a scalable and data-driven methodology for categorizing OGD impact areas compared to previous research that relies on predefined classifications or manual taxonomies. However, the study has limitations, including a relatively small dataset, brief use cases, and the inherent subjectivity of taxonomic classification, which requires further validation by domain experts. This research contributes to the systematic assessment of OGD initiatives and provides a foundational framework for policymakers and researchers aiming to maximize the benefits of open data.
Ključne besede: open government data, topic modeling, taxonomy development, machine learning
Objavljeno v DKUM: 28.08.2025; Ogledov: 0; Prenosov: 3
Celotno besedilo (600,19 KB)
Gradivo ima več datotek! Več...