Enhancing the learning process of people with special needs using wearables and Artificial Intelligence

Bakthavatchaalam, Venkat

Abstract

Wearables are widely used in monitoring various physical activities and biological signals of an individual to improve their current lifestyle. This research proposes a novel way to enhance the learning of individuals with special needs, by collecting holistic data through wearables and analysing through Machine Learning (ML) models. The models can then be used not only to identify and understand the gamut of factors that influence their learning, but also to enhance their learning processes overall.
This research explores a more holistic (Physiological, Psychological and Technical) picture by using a plurality of sensors that measures both their physical and emotional states. The wearable is equipped with sensors measuring their skin temperature and conductivity, muscle activities, heart rate and motion sensors. These data points can be visualised using a Mobile App in real time, thus making it easy and quick to view data on their anxiety and heart rate levels etc. during different learning process. The data collected from the wearable can be analysed to form models using machine learning algorithm such as K-Nearest Neighbours and Multi-Layer Perceptron (MLP). This data driven process would provide a deep insight of the mutual interplay of both physical and emotional states influencing their learning processes and to recommend and enhance developmental strategies.
United Nation’s Sustainable Development Goal no. 4 strongly advocates inclusivity and to leave no one behind. This research seeks to contribute to that goal by attempting to understand and enhance the learning process of individuals with special needs. Thus, contributing to the theme and the spirit of MLW.

• Presentation Conference Type: Conference Paper (unpublished)
• Conference Name: UNESCO: Mobile Learning Week 2019
• Start Date: Mar 4, 2019
• End Date: Mar 8, 2019
• Deposit Date: Aug 5, 2021
• Series Title: UNESCO: MLW
• Public URL: https://uwe-repository.worktribe.com/output/6550441