Research & Publications

Academic work in adaptive learning, serious games, and intelligent tutoring systems

Thesis

Adaptive Instructional System Design for Disaster Preparedness: A 13-Model Bayesian Knowledge Tracing Approach in Serious Games

In progress

Kaçar, S.2026

This thesis presents AfetAkademi, an adaptive serious game for disaster preparedness education targeting children aged 8–16. The system integrates a 13-model BKT architecture with Piaget-aligned age adaptation, Socratic AI mentoring, and real-time learning analytics.

Manuscripts in Preparation

AfetAkademi: A 13-Model Bayesian Knowledge Tracing System for Adaptive Disaster Education in Children

Manuscript in preparation

Kaçar, S. (2026)

Target venue: AIED 2026 (International Conference on Artificial Intelligence in Education)

Abstract: This paper presents a 13-model BKT architecture designed for a serious game targeting disaster preparedness education. The system extends standard BKT with 10 specialist models — including MetaCognition, Confidence, ErrorPattern, and HelpSeeking — indexed per objective, and calibrates parameters using Piaget's developmental stages across three age groups (8–10, 11–13, 14–16).

Piaget-Aligned BKT Parameter Calibration in Game-Based Adaptive Learning Systems

Manuscript in preparation

Kaçar, S. (2026)

Target venue: EDM 2026 (Educational Data Mining)

Technical Writing

Talks & Presentations

No talks scheduled yet.

Interested in a collaboration or speaking invitation? info@sibelkacar.com

Conference & Accepted Submissions

Afet Akademi: Dijital ve Adaptif Afet Eğitimi Modeli

Accepted

Graduate Teacher Studies Congress (LOCK) — Disaster Education, 14–17 May 2026

Afet Akademi is a digital, adaptive disaster education platform designed with the ADDIE model. It aims to increase disaster awareness through gamification while positioning the teacher as a data-informed learning designer.

The platform delivers 48 learning outcomes via quizzes, matching tasks, scenarios, strategy and time-based games. An SM-2–based adaptive engine identifies students’ strengths and weaknesses, manages review schedules, and supports teachers with anonymized learning analytics rather than grades or ranking.

Afet Akademi is a digital, adaptive disaster education platform designed using the ADDIE instructional design model. Its goal is to increase disaster awareness through gamification while repositioning the teacher as a data-informed learning designer.

Purpose

Although disaster education is critical for fostering safe behavior and social solidarity, traditional approaches often fail to reach all learners equitably due to limitations in time, materials, and teacher workload. This work presents Afet Akademi as a scalable, learning-analytics–driven digital platform that aims to enhance disaster awareness through gamification and support teachers in making evidence-based instructional decisions.

Target Audience

Primary, lower secondary, and upper secondary students; teachers and academics; and educators who wish to integrate disaster education into their courses.

Overview of the Practice

Afet Akademi (www.afetakademi.com.tr) delivers 48 learning outcomes via quizzes, matching tasks, scenarios, strategy games, reading activities, and time-based games. Teachers enroll students using class codes and monitor progress through anonymized dashboards. An SM-2–based adaptive engine estimates students’ mastery for each outcome, recommends suitable game types for weaker areas, and schedules repetitions. The adaptive structure supports growth without relying on grades or ranking, and makes the “From Past to Future Teacher” theme concrete by shifting the teacher’s role from content delivery to the design of learning experiences. The platform is browser-based, free to use, and aligned with data protection regulations through anonymized data collection.

Outcomes and Implications

The model offers a repeatable digital framework that aims to foster safe behavior and disaster awareness among students. Teachers can track overall class progress and weak outcomes anonymously, and redesign subsequent lessons with targeted games or discussion-based activities when many students struggle with a specific outcome. The scalable digital structure is intended to reduce teacher workload, strengthen instructional processes through learning analytics, and contribute to digital equity in education. Future research will focus on classroom impact studies examining learning gains and behavioral outcomes associated with data-informed, game-based disaster education.

Live Research Platform

afetakademi.com.tr

Learning Analytics & XAI Platform — 7 game types, 48 learning outcomes, KVKK/GDPR compliant anonymous data collection, CSV export for SPSS / R / Python