Gated Attention Based Deep Learning Model for Analyzing the Influence of Social Media on Education

Abstract

ABSTRACT The major aim of education organisation is to ensure better education to students and reducing the failure percentage of poor students. Early prediction of student’s performance is a complex process to enhance the academic performance. However, it is complex to analyse huge data manually. Therefore, an automated model is essential for mining the educational performance of the students. Educational data mining (EDM) is more efficient when compared with statistical approaches utilised for exploring data for predicting the performance of students. This work has four major processes: (i) pre-processing, (ii) feature extraction, (iii) classification and (iv) evaluating performance. Initially, the data are pre-processed and the necessary features are extracted using Word2Vec and Term Frequency & Inverse Gravity Moment. The major novelty of this work is to utilise deep learning model attention based on bi-directional long short-term memory convolutional neural network with gate mechanism and to find the students’ academic performance. The performance of the proposed model is compared over the metrics like accuracy, sensitivity, precision and F-measure, respectively. Results prove that the proposed model highly influences the performance of students. The analysis showed that the students who utilise social networks for the whole week showed less performance when compared to students who utilise social networks only during weekends.