Machine Learning-Based Classification of Mental Health Status on Social Media: A Case Study of Kunduz, Afghanistan

Аннотация

The proliferation of user-generated content results from social media's rapid development. An interdisciplinary field called computational cyber- psychology uses machine learning techniques to investigate fundamental psychological tendencies. Our study uses social media usage patterns to infer users' mental health status. The emergence of social media platforms in Afghanistan has profoundly affected the country's young people. This study aims to implement a predictive model using three machine-learning algorithms. The three selected algorithms are Naïve Bayes, Random Forest, and K-Nearest Neighbor (K-NN). The dataset contained in this study was collected through a questionnaire from students at Kunduz public and private universities in 2024. Data preprocessing is done before implementing the predictive models. The dataset is prepared carefully to ensure well-balanced samples in each category. The study reveals that Random Forest is the best classifier, with an accuracy of 77%. The study directly benefits the other researchers and policy or decision-makers in Afghanistan. Keywords: Random Forest, K-NN, Social Media, Mental Health, Afghanistan. European International Journal of Multidisciplinary Research 6 https://eipublication.com/index.php/eijmrms and Management Studies European International Journal of Multidisciplinary Research and Management Studies Introduction: A notable aspect of the Internet's dataset into training and test sets. One or more explosive growth has been the creation of social classification algorithms run through the training set, media, which has drawn a large number of people and the classifier models are subsequently developed. eager to express themselves on sites like Facebook, The test set is then used to assess the accuracy of the Twitter, and others. Users are creating a cyberspace models [3]. Previous studies show that the dataset and that interacts with and reflects the actual world application have a major impact on the accuracy and through their online habits. As a result, a person's efficiency of machine learning algorithms. For example, online behavior may be a good reflection of their algorithms like Support Vector Machines (SVM) have psychological traits offline. Due to demands, the demonstrated efficacy in mental health prediction outside world, and other factors, an increasing number research because of their capacity to handle high- of people these days are experiencing mental illnesses dimensional data, even if Random Forest has including depression, anxiety, stress, and so on. These demonstrated strong results in a variety of prediction mental illnesses can have a serious negative impact on tasks. A fair evaluation is made possible by using a range users' lives and can even cause suicides. In the past, of classifiers, which also offers insights into which individuals with mental health issues would have been algorithms work best in certain mental health prediction told to see a therapist or could have gone out on their settings [4]. This study's inclusion of Naïve Bayes, K-NN, own to get psychotherapy. Individuals become aware and Random Forest serves as a comparative basis to of mental health issues through surveys or gut feelings. identify the optimal classifier for social media-based Lack of resources may prevent psychotherapy from mental health prediction. Subsections A to C briefly becoming accessible, even in cases where a mental introduce the classifiers selected for this study, while health issue is identified. These days, the Internet Section III describes methodology of the work. offers a fresh way to handle such circumstances [1]. A. Naïve Bayes: In Naïve Bayes learning, a Bayesian Due to increased social comparison and validation- probabilistic model accredits a back-class probability to seeking tendencies, people who use social media an instance P(Y=Yj|X=Xi). The simple Naïve Bayes regularly are Due to increased social comparison and algorithm uses these probabilities to accredit an validation-seeking tendencies, people who use social example to a class. Naive Bayes classifier converges media regularly are more likely to suffer from mental faster than logistic regression, so it requires only less health conditions like anxiety, depression, and training data. This method is prevalent for different loneliness, according to studies. Social media usage applications for several features. This classifier can be frequently reflects offline habits, which can be used to trained with reasonable accuracy in a supervised create mental health assessment prediction models learning setting, and its performance is satisfactory in [2]. Research from the World Health Organization in many complex real-life problem situations. Naive Bayes 2018 found two million Afghans struggling with mental is technically less precise than other classifiers and can distress, and these numbers are likely much higher result in a higher lower rate [5]. today. Still, many suffer in silence. Th