AI educational mobile app using deep learning approach.

Moving to Industrial Revolution (IR 4.0), the early education sector is not left behind. More of the teaching method is being digitized into a mobile application to assist and enhance the children’s understanding. On the other hand, most of the applications offer passive learning, in which the child...

Full description

Saved in:
Bibliographic Details
Main Authors: Mohd. Nasir, Haslinah, Brahin, Noor Mohd. Ariff, Ariffin, Farees Ezwan Mohd Sani, Mispan, Mohd. Syafiq, Abd. Wahab, Nur Haliza
Format: Article
Language:English
Published: Politeknik Negeri Padang 2023
Subjects:
Online Access:http://eprints.utm.my/105943/1/NurHalizaAbdWahab2023_AiEducationalMobileAppUsingDeepLearningApproach.pdf
http://eprints.utm.my/105943/
https://joiv.org/index.php/joiv/article/view/1247/742
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Moving to Industrial Revolution (IR 4.0), the early education sector is not left behind. More of the teaching method is being digitized into a mobile application to assist and enhance the children’s understanding. On the other hand, most of the applications offer passive learning, in which the children complete the activity without interacting with the environment. This study presents an educational mobile application that uses a deep learning approach for interactive learning to enhance English and Arabic vocabulary. Android Studio software and Tensorflow tool were used for this application development. The convolution neural network (CNN) approach was used to classify the item of each category of vocab through image recognition. More than thousands of images each time were pre-trained for image classification. The application will pronounce the requested item. Then, the children will need to move around looking for the item. Once the item’s found, the children must capture the image through the camera’s phone for image detection. This approach can be integrated with teaching and learning techniques for fun learning through interactive smartphone applications. This study attained high accuracy of more than 90% for image classification. In addition, it helps to attract the children's interest during the teaching using the current technology but with the concept of ‘Play’ and ‘Learn’. In the future, this paper recommended the involvement of IoT platforms to provide widen applications.