Original research
par
Garg, Sanjana et al
Date de publication
2021
Géographie
USA
Langue de la ressource
English
Texte disponible en version intégrale
Oui
Open Access / OK to Reproduce
Oui
Évalué par des pairs
Yes
L’objectif
Because those who misuse fentanyl are often a hidden and high risk group, innovative methods for identifying individuals at risk for fentanyl misuse are needed. Machine learning has been used in the past to investigate discussions surrounding substance use on Reddit, and this study leverages similar techniques to identify risky content from discussions of fentanyl on this platform.
Constatations/points à retenir
Our machine learning risk model was able to detect posts or comments labeled as risky by our clinical experts with 76% accuracy and 76% sensitivity. Furthermore, we provide a vocabulary of community-specific, colloquial words for fentanyl and its analogues.
La conception ou méthodologie de recherche
A codebook was developed by clinical domain experts with 12 categories indicative of fentanyl misuse risk, and this was used to manually label 391 Reddit posts and comments. Using this data we built machine learning classification models to identify fentanyl risk.
Mots clés
Harm reduction
Illegal drugs
Digital health