Traditional Machine Learning Models and Bidirectional Encoder Representations From Transformer (BERT)-Based Automatic Classification of Tweets About Eating Disorders: Algorithm Development and Validation Study
Authors:
José Alberto Benítez-Andrades,
José-Manuel Alija-Pérez,
Maria-Esther Vidal,
Rafael Pastor-Vargas,
María Teresa García-Ordás
Abstract:
Background: Eating disorders are increasingly prevalent, and social networks offer valuable information.
Objective: Our goal was to identify efficient machine learning models for categorizing tweets related to eating disorders.
Methods: Over three months, we collected tweets about eating disorders. A 2,000-tweet subset was labeled for: (1) being written by individuals with eating disorders, (2…
▽ More
Background: Eating disorders are increasingly prevalent, and social networks offer valuable information.
Objective: Our goal was to identify efficient machine learning models for categorizing tweets related to eating disorders.
Methods: Over three months, we collected tweets about eating disorders. A 2,000-tweet subset was labeled for: (1) being written by individuals with eating disorders, (2) promoting eating disorders, (3) informativeness, and (4) scientific content. Both traditional machine learning and deep learning models were employed for classification, assessing accuracy, F1 score, and computational time.
Results: From 1,058,957 collected tweets, transformer-based bidirectional encoder representations achieved the highest F1 scores (71.1%-86.4%) across all four categories.
Conclusions: Transformer-based models outperform traditional techniques in classifying eating disorder-related tweets, though they require more computational resources.
△ Less
Submitted 8 February, 2024;
originally announced February 2024.
