Companion for elderly: Chatbot and mini-games for Alzheimer’s prediction

Tan, Qi Yang (2025) Companion for elderly: Chatbot and mini-games for Alzheimer’s prediction. Final Year Project, UTAR.

PDF Download (3539Kb)

Abstract

In recent years, many elderly people who are living alone have faced lots of difficulties, such as a lack of companionship, difficulty managing daily tasks, and unawareness of health issues like Alzheimer's disease. Although there are a few mobile applications available to help solve these issues, they often lack a comprehensive, all-in-one solution tailored specifically for the elderly. This project aims to study the development of a mobile application designed to help elderly people by combining multiple features into a single platform. The proposed application addresses limitations found in existing mobile applications such as CogniFit, Replika, and MyTheraphy which they do not contain the features cognitive assessment, companionship feature, and tools to assist in daily lives in a single platform. Thus, the proposed application will provide an integrated solution to assist the elderly in their daily lives. It includes a chatbot that serves as a virtual companion to engage users in meaningful conversations and a minigame-based cognitive assessment tool which reference from Mini-Mental State Examination (MMSE) test to help monitor cognitive health. The polynomial regression model with average MAE score of 0.639, average RMSE score of 0.915, and average R² score of 0.595 is integrated as predictive model for Alzheimer’s disease. The application is developed using Android Studio, Unity Game Engine with C#, and Android SDK with Kotlin and JavaScript, and is designed to be compatible with Android smartphones. Besides, users can use the utility tools built in the application such as the contact and the reminder to support in their daily activities.

Actions (login required)