"Analyzing the Influence of Social Media on Students' Academic Performance Using Machine Learning and Deep Learning Approaches "

Date
2023
Editor
Journal Title
Volume
Issue
Journal Title
Journal ISSN
Volume Title
Abstract
Analyzing the Influence of Social Media on Students' Academic Performance Using Machine Learning and Deep Learning Approaches : Social media (SM) are online media technologies that allow people to share and exchange information, ideologies, preferences, and other forms of expression. Since the 1990s, social media platforms have been utilized as an effective method of communication. People can more readily engage with one another and share ideas, information, and opinions because of the proliferation of social media platforms. The worldwide community is more connected than it has ever been before because of the rise of social media. Even in Bangladesh, social media users are increasing rapidly. Studies have shown that students' academic performance tends to worsen when they spend more time on social media. The reason for this is that they choose to engage in conversation with their friends on social networking sites rather than read a book during their downtime. Their social lives and mental health are also negatively impacted, in addition to their academic performance. On the other hand, utilizing social media does come with a few advantages to consider as well. Currently, social media is responsible for the creation of a great deal of both negative and positive aspects. Students are placing a greater emphasis on their use of social media than they are on their ability to read and write, which is leading to a significant loss of both time and academic performance. Consequently, the purpose of this study is to investigate the impact of students' use of social media on their academic performance by applying machine learning (ML) and deep learning (DL) algorithms. As there have already been many studies done in this field, our study focused on university students in Bangladesh. This study used a unique set of data. In this study, we develop this model by utilizing both traditional ML and DL techniques. These algorithms are referred to as Random Forest (RF), Decision Tree (DT), K-Nearest Neighbor (KNN), Gaussian Naïve Bayes (GNB), and Mini Batch Gradient Descent (MBGD). Accuracy is used as the evaluation metric of our models. Among them, Mini Batch Gradient Descent achieved the highest accuracy rate of 99.25% and Random Forest achieved an accuracy rate of 96.43%. This study, on the whole, came up with satisfactory results, successfully forecasting with an excellent level of accuracy.
Description
Keywords
Citation
Department Name
Electrical and Computer Engineering
Publisher
North South University
Printed Thesis
DOI
ISSN
ISBN