cover
Contact Name
Dwi Agus Kurniawan
Contact Email
dwiagus.k@unja.ac.id
Phone
+6282380245589
Journal Mail Official
jiituj@unja.ac.id
Editorial Address
Jalan Raya Jambi-Ma.Bulian, KM.15 Mendalo Indah
Location
Kota jambi,
Jambi
INDONESIA
Jurnal Ilmiah Ilmu Terapan Universitas Jambi
Published by Universitas Jambi
ISSN : 25802240     EISSN : 25802259     DOI : https://doi.org/10.22437/jiituj.v6i2
JIITUJ publish the result of research on applied science and education (Research of applied science and education) such as: the research result on applied science and education such as curriculum development and learning, character education, technology and instructional innovation, and learning evaluation. the research result on applied science and technology such as the development of applied technology and applied arts, appropriate technology, designing information systems, the research result on applied science and economic development the research result on applied science and public health. JIITUJ is a double-blind peer-reviewed journal, published three (3) times a year by Research Institutions and Community Service (LPPM), Universitas Jambi, Indonesia. JIITUJ is open to academic circles and university researchers, research institutes, librarians, graduate and postgraduate students to share information on research results.
Articles 512 Documents
FORECASTING INDONESIAN INFLATION: A RECURRENT NEURAL NETWORK-BASED DEEP LEARNING APPROACH Adyatma, Wimala Eka; Windhani, Kikin
Jurnal Ilmiah Ilmu Terapan Universitas Jambi Vol. 10 No. 5 (2026): Volume 10, Nomor 5, October 2026
Publisher : LPPM Universitas Jambi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.22437/jiituj.v10i5.53969

Abstract

Forecasting models in the economic field have traditionally relied on conventional statistical methods. However, these methods often face limitations when dealing with volatile and non-stationary data, such as inflation in Indonesia. A deep learning approach based on Recurrent Neural Networks (RNN) is proposed as an alternative for forecasting Indonesian inflation. This study aims to evaluate the accuracy of the RNN method as a forecasting model for inflation in Indonesia. The study employs a quantitative time-series forecasting design using secondary monthly inflation data from January 2003 to January 2026, collected from Bank Indonesia. The analytical framework uses an RNN-based deep learning model with hyperparameter optimization, Min-Max Scaling, and a chronological 70:30 train-test split. Model performance is evaluated using Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE). The results indicate that the optimal forecasting performance is achieved using the Tanh activation function, 10 RNN units, and a dropout rate of 0.05, with an RMSE of 0.369869 and an MAE of 0.266443 on the test data. The RNN model projects a stable inflation trend ranging from 3.38% to 3.9475%. These findings suggest that RNN can serve as a complementary approach for forecasting Indonesian inflation.
CRISIS COMMUNICATION IN MANAGING THE REPUTATION OF MARINE TOURISM DESTINATIONS DUE TO WASTE PROBLEMS: A CASE STUDY OF JARING HALUS VILLAGE, LANGKAT REGENCY Faustyna, Faustyna; Rudianto, Rudianto; Santoso, Puji; Hendra, Yan
Jurnal Ilmiah Ilmu Terapan Universitas Jambi Vol. 10 No. 5 (2026): Volume 10, Nomor 5, October 2026
Publisher : LPPM Universitas Jambi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.22437/jiituj.v10i5.59885

Abstract

The marine debris crisis confronting Jaring Halus, a community-based mangrove and marine tourism village in Secanggang District, Langkat Regency, North Sumatra, threatens not only its ecological condition but also its reputation as a tourism destination; this study analyzes the crisis communication strategies applied to manage that reputation, using Situational Crisis Communication Theory and Image Restoration Theory. Using a qualitative case study design, the study collected secondary data through document analysis and non-participant observation of digital content, news coverage, social media discourse, and government records, and analyzed them using Miles' interactive technique. The waste problem in Jaring Halus is predominantly a victim cluster crisis, since most debris originates from land-based and transboundary sources carried by the Malacca Strait current rather than the local community; however, media framing still places reputational risk on the destination. Effective responses combine rebuilding strategies (community-based waste banks and mangrove rehabilitation), bolstering strategies rooted in local wisdom, particularly the jamu laut ritual and mangrove biodiversity narratives, and diminishment messaging clarifying the transboundary origin of the debris. These findings extend crisis communication and image restoration research, previously focused mainly on large-scale disasters and internally caused crises, to an externally caused, small-scale, community-managed destination; the novelty lies in the resulting integrated multi-stakeholder crisis communication model combining diminishment, rebuilding, and bolstering strategies for environmentally caused reputational threats, offering practical implications for village governments, tourism agencies, and community tourism groups in designing evidence-based, culturally grounded communication responses.