Marzuqi, Ahmad and Caraka, Rezzy Eko and Gio, Prana Ugiana and Chen, Rung Ching and Noh, Maengseok and Pardamean, Bens (2024) Unlocking insights from Ministry of Marine Affairs and Fisheries annual reports using LDA: a deep dive into SDG 14. TELKOMNIKA (Telecommunication Computing Electronics and Control), 22 (4). p. 784. ISSN 1693-6930
Full text not available from this repository. (Request a copy)Abstract
Annual reports serve as vital instruments for government ministries and agencies, enabling transparency and accountability in managing state budgets (APBN) and activities, thereby fulfilling a crucial role in public accountability, particularly in the context of sustainable development goal (SDG) 14. However, due to their extensive nature, it becomes imperative to conduct topic modeling analysis to discern trends and topics within these reports. In this study, latent Dirichlet allocation (LDA), a prominent topic modeling technique, is employed to analyze the annual reports of the Ministry of Marine Affairs and Fisheries (KKP) Indonesia from 2015 to 2022. Utilizing the coherence score as an evaluation metric, we assess the quality of topic models across each report year. Our findings underscore the consistent emphasis on fisheries and marine-related initiatives, emphasizing their relevance to SDG 14 and Indonesia’s maritime landscape. Ultimately, this study offers valuable insights to inform strategic planning and decision-making processes within the KKP, contributing to the advancement of SDG 14 and promoting sustainable development in Indonesia’s fisheries and marine sectors.
| Item Type: | Article |
|---|---|
| Uncontrolled Keywords: | Annual reports, Fisheries, Latent Dirichlet allocation, Marine resources, Ministry of Marine Affairs and Fisheries, Strategic planning, Topic modeling |
| Subjects: | Ocean Sciences & Technology Government Inventions for Licensing |
| Depositing User: | Rizzal Rosiyan |
| Date Deposited: | 26 Nov 2025 06:29 |
| Last Modified: | 26 Nov 2025 06:29 |
| URI: | https://karya.brin.go.id/id/eprint/55307 |


Dimensions
Dimensions