SMARTS

SMARTS analyzes Reddit and Twitter posts to extract and characterize substance use behaviors and to predict individuals receptive to addiction recovery interventions for epidemiological and intervention research.

Key Features:

• Data Source Utilization: Harvests user-generated content from Reddit and Twitter to obtain real-time information on substance use behaviors.
• Natural Language Processing and Machine Learning: Applies NLP and machine learning techniques to analyze posts related to opioids, cannabis, kratom, alcohol, and cigarettes.
• Predictive Modeling: Implements predictive models to identify individuals who may be receptive to addiction recovery interventions.
• Data Extraction and Analysis: Extracts post topics, sentiment, drug- and recovery-related term usage, geolocation data, and age demographics from social media posts.
• Visualization Tools: Provides visualization capabilities to interpret trends and patterns in substance use data.
• Reference Data: Includes reference distributions derived from a dataset of 170,097 drug users for contextualizing findings.

Scientific Applications:

• Epidemiological Surveillance: Enables analysis of substance abuse epidemiology at individual and community levels using social media-derived data.
• Intervention Targeting: Supports design and targeting of addiction recovery interventions by predicting receptive individuals.
• Trend and Pattern Analysis: Facilitates characterization of temporal and demographic patterns in use of opioids, cannabis, kratom, alcohol, and cigarettes.

Methodology:

Integrates NLP and machine learning algorithms, including sentiment analysis, to process and analyze Reddit and Twitter posts and applies predictive modeling to identify individuals receptive to recovery interventions.