Abstract
Depression is a growing global issue that requires urgent attention, particularly in Thailand. Without timely intervention, it can lead to serious consequences such as health complications, reduced social participation, and even suicide. Consequently, there has been increased research focused on developing algorithms to identify and classify social media data according to depression severity. This study aimed to evaluate the extent to which such algorithms are a suitable tool for the preliminary classification of Thai Facebook text into different levels of depression severity. The methodology involved data collection through surveys and Facebook scraping, followed by data preprocessing, feature engineering, and model training and validation. The models employed included Support Vector Machine, Logistic Regression, Decision Tree, and K-means clustering, applied at both subject- and text- levels. Labels were derived from the Thai Depression Inventory and expert evaluations. The results demonstrated that machine learning algorithms can serve as effective tools for classifying Facebook text into varying levels of depression severity. However, challenges such as class imbalance, limited data availability, and discrepancies between self-assessment results and online behavior indicate that further research is required to enhance model accuracy and reliability before these tools can be effectively applied in real-world settings.
DOI
10.56808/2673-060X.5758
Recommended Citation
Yoneyama, Keito
()
"Assessing The Reliability Of Machine Learning Models In Classifying Depression Severity Using Facebook Text,"
Chulalongkorn Medical Journal: Vol. 70:
Iss.
5, Article 1.
DOI: https://doi.org/10.56808/2673-060X.5758
Available at:
https://digital.car.chula.ac.th/clmjournal/vol70/iss5/1