Stress mental health symptom assessment mobile application for young adults

Lee, Chun Hoong (2023) Stress mental health symptom assessment mobile application for young adults. Final Year Project, UTAR.

Preview

PDF Download (10Mb) | Preview

Abstract

Mobile health applications, better known as “mHealth” applications, are getting popular nowadays. Mobile digital health technology enhances patient care by improving condition monitoring and diagnosis methods, resulting in more timely and comprehensive care. In this fast-paced world, young adults are getting stressed compared to previous years as they experience more difficulties than before in terms of work, school, and relationships. Stress harms health and productivity, whether a constant struggle or an occasional flare-up. This project is to study the development of a stress mental health symptom assessment mobile application to help people, especially young adults or university students, measuring their stress levels and then carry out appropriate activities for relieving stress. The proposed application intends to solve several limitations found in the existing stress management applications in the mobile applications market. Some existing applications such as Headspace and Smilin Mind do not have a proper stress level assessment backed by solid scientific studies. Therefore, this proposed stress mental health symptom assessment mobile application can allow users to conduct a stress assessment session in the application to generate a report of the stress level. It is developed using Android Studio, React Native, Android SDK, Javascript, Google Firebase, and an Android smartphone. A series of practical activities are recommended to the users to relieve the stress. Moreover, users can provide more inputs based on their feelings throughout the day. One of the primary functionalities of the application is to incorporate a machine learning algorithm which is K-Nearest Neighbor (KNN) classification technique for panic attack prediction feature to enhance the emotional identification and offering users an artificial intelligence (AI) chatbot. As for the panic attack prediction feature, the application will transmit user-input responses to a model previously trained using historical data. Before this stage, the KNN algorithm is employed to construct the model using Google Form response data as part of the application development process. This model will then be integrated into the React Native application for use. It will gather information such as the user’s gender, age, current course of study, current year of study, marital status, and any previous instances of seeking special treatment. Subsequently, employing the K-Nearest Neighbors (KNN) algorithm, the model shall forecast the likelihood of experiencing a future panic attack. Remarkably, the KNN model exhibits a testing accuracy of 70.37%, signifying a commendable outcome from both the training and testing phases. This information can empower them to take necessary steps for preparation or prevention if a panic attack is anticipated in the future.

Actions (login required)