Serious Game Peringatan Dini Banjir Berbasis Multicriteria Decision Making Menggunakan Sistem Rekomendasi Prediktif

Rizal, Muh (2026) Serious Game Peringatan Dini Banjir Berbasis Multicriteria Decision Making Menggunakan Sistem Rekomendasi Prediktif. Doctoral thesis, Institut Teknologi Sepuluh Nopember.

Text 7022211021-Doctoral.pdf - Accepted Version Restricted to Repository staff only Download (9MB) | Request a copy

Abstract

Banjir merupakan salah satu bencana alam yang paling sering terjadi dan menyebabkan kerugian yang signifikan dalam aspek ekonomi, sosial, dan lingkungan. Kompleksitas faktor-faktor penyebab banjir, mulai dari intensitas curah hujan, perencanaan penggunaan lahan, kapasitas drainase, hingga perubahan iklim, membutuhkan adanya sistem prediksi yang lebih akurat serta mekanisme edukasi yang mampu meningkatkan kesiapsiagaan masyarakat. Penelitian ini bertujuan untuk mengembangkan platform terintegrasi berupa Flood Alert Serious Game berbasis Multi-Kriteria Decision Making Pendekatan yang digunakan mencakup tiga komponen inti. Pertama, sistem prediksi banjir dikembangkan menggunakan algoritma Machine Learning dengan fokus pada Random Forest and Support Vector Machine (SVM) yang dioptimalkan dengan metode pemilihan fitur Quasi-Oppositional Biogeography-Based Optimization Whale Optimization (QOBL-BWO). Hasil penelitian menunjukkan bahwa model mampu mencapai akurasi prediksi lebih dari 99% dengan konsistensi tinggi dalam metrik evaluasi lainnya. Kedua, modul MCDM (TOPSIS) digunakan untuk mengubah output model prediksi menjadi status siaga banjir yang dapat ditindaklanjuti, berdasarkan berbagai kriteria seperti curah hujan, ketinggian air, debit sungai, pasang surut, suhu, dan kelembaban. Ketiga, game interaktif serius dirancang sebagai media edukasi yang mensimulasikan skenario banjir nyata, mengintegrasikan output prediksi dan MCDM ke dalam alur permainan, dan menyertakan sistem rekomendasi berbasis prediksi untuk memberikan saran tindakan mitigasi yang relevan dan kontekstual bagi para pemain. Hasil penelitian menunjukkan bahwa integrasi ketiga komponen tersebut menghasilkan platform yang tidak hanya akurat dalam memprediksi banjir, tetapi juga efektif dalam meningkatkan literasi bencana, persepsi risiko, dan kepercayaan pengguna dalam mengambil keputusan. Pengujian pengguna menunjukkan peningkatan yang signifikan dalam pemahaman strategi mitigasi setelah berinteraksi dengan game. Dengan demikian, penelitian ini berkontribusi pada pengembangan sistem peringatan banjir generasi baru yang tidak hanya berbasis analisis prediktif, tetapi juga memanfaatkan pendekatan gamifikasi untuk mendukung kesiapsiagaan masyarakat.
===================================================================================================================================
Floods are one of the most frequent natural disasters and cause significant losses to economic, social, and environmental aspects. The complexity of flood-causing factors, ranging from rainfall intensity, land use, drainage capacity, to climate change, demands a more accurate prediction system and an education mechanism that can improve community preparedness. This research aims to develop an integrated platform in the form of Flood Alert Serious Game based on Multi-Criteria Decision Making
The approach used includes three core components. First, the flood prediction system was developed using a Machine Learning algorithm with a focus on Random Forest and Support Vector Machine (SVM) which was optimized with a feature selection method based on Quasi-Oppositional Biogeography-Based Optimization Whale Optimization (QOBL-BWO). The results showed that the model was able to achieve prediction Accuracy of more than 99% with high consistency on other evaluation metrics. Second, the MCDM module (TOPSIS) is used to convert the output of the prediction model into an actionable flood Alert status, based on various criteria such as rainfall, water level level, river discharge, tide, temperature, and humidity. Third, the interactive Serious Game is designed as an educational medium that simulates real flood scenarios, integrates prediction outputs and MCDM into the gameplay, and includes a prediction-based recommendation system to provide relevant and contextual suggestions for mitigation actions for players. The results show that the integration of the three components results in a platform that is not only accurate in predicting floods, but also effective in improving disaster literacy, risk perception, and user confidence in decision-making. User tests showed a significant improvement in understanding mitigation strategies after interacting with the game. Thus, this research contributes to the development of a new generation of flood Warning systems that are not only based on prediction analysis, but also utilize gamification approaches to support community preparedness.

Item Type:	Thesis (Doctoral)
Uncontrolled Keywords:	banjir, serious game, prediksi, multicriteria decision making ============================================================ flood, serious game, prediction, multicriteria decision Making
Subjects:	T Technology > T Technology (General) > T385 Visualization--Technique
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20001-(S3) PhD Thesis
Depositing User:	Muh. Rizal H
Date Deposited:	23 Jan 2026 02:14
Last Modified:	23 Jan 2026 02:14
URI:	http://repository.its.ac.id/id/eprint/130180

Actions (login required)

View Item