top of page

Project ID:

SCEN04302

SUSTAIN-AI Malaysia: An AI-Powered Decision Support System for Sustainable Tourism Pressure Analysis in Malaysia

Project Title:

Category:

Science and Engineering

Inventors:

Tey Jay Yang, GAN WEI XING (MEMBER), CHANG JING XUAN (MEMBER), NGUI YU JIE (MEMBER), DR. LIM WEI JER (ADVISOR)

Institution/Company:

Southern University College

Invention Description/ Abstract:

SUSTAIN-AI Malaysia is an AI-powered decision support platform designed to support sustainable tourism planning through official Malaysian data. It integrates an interactive website, Power BI dashboard, regional tourism demand indicators, machine learning analysis, and an AI evidence assistant called TAPi. Users can explore tourism patterns across Malaysian states, compare regional demand, examine model results, and understand the evidence behind each finding. Its key innovation is combining transparent predictive modelling with practical decision support. The platform evaluates multiple machine learning models, reports prediction errors, and clearly distinguishes observed data from experimental projections. Its Decision Lab transforms regional findings into evidence-based investigation priorities, while the Evidence Centre documents data sources, methods, and limitations. Guided by the framework “Observe. Predict. Explain. Act.”, SUSTAIN-AI Malaysia aims to improve tourism data accessibility, encourage responsible AI adoption, and help stakeholders make more informed decisions for sustainable tourism development in Malaysia.

Invention Technical Description

SUSTAIN-AI Malaysia integrates official Malaysian tourism datasets, statistical processing, machine learning, and interactive visualisation within a software-based decision support system. Its regional demand indicator combines normalised visitor intensity, overnight tourism intensity, and annual visitor growth using declared weights of 50%, 30%, and 20%. The predictive component analyses domestic tourism data from 16 Malaysian regions between 2018 and 2025. Five models are evaluated: Last-Year Naive, Ridge Regression, Random Forest, Extra Trees, and XGBoost. Temporal validation uses 2022–2023 for model selection and 2024–2025 for later assessment. The selected Last-Year Naive model achieved a WMAPE of 13.87%. Results are presented through an interactive website and Power BI dashboard. TAPi provides AI-assisted explanations, while the Evidence Centre documents data provenance, calculation methods, and limitations. The system distinguishes observed statistics, derived indicators, and experimental projections to support transparent, evidence-based tourism analysis.

Demostration/ Presentation Video

Poster/ Broucher/ Invention Photo

Additional Documents

> Photo of the Invention
> Additional Technical Document
> Additional Photo of the Invention

Copyright ©2026 by I-RICE 2026. All rights reserved

bottom of page