Identifying personality traits of WhatsApp users based on frequently used emojis using deep learning.

Social networks have become an element of individual life in the present era and have many applications in economic, social, commercial, and even educational and academic fields and aspects. Daily use of various messaging applications affects people's lifestyles, and the usage method of such applications differs depending on the user's personality. WhatsApp messenger is one of the most popular social media applications widely used by people enabling them to send messages in text, voice, image, and emojis. In this paper, a Long Short-Term Memory (LSTM) neural network is designed to identify the personality trait of WhatsApp users by analyzing the most frequently used emojis. The users are classified into 16 classes involving four psychological aspects considered as introverted or extroverted, happy or depressed, optimistic or pessimistic, and neurotic or calm. For this purpose, 13,688 samples were collected from volunteer users, including screenshots of frequently used emojis and their beliefs about mentioned feelings. The screenshot images were converted to numerical codes using image processing, and the proposed LSTM is implemented in MATLAB 2021b using the deep learning toolbox. The simulation results show that the proposed network is trained with 96.3% accuracy and can predict the personality type of individuals with 95.48% accuracy based on users' most frequently used emojis. The proposed method also achieves a minimum 93.99% score in the Precision, Recall, and F-measure criteria. The comparison of the results obtained by the Random Forest algorithm shows the superiority of the proposed model. [ABSTRACT FROM AUTHOR]