Researchbertrecoveryphoenix appsentiment analysis
Phoenix App Shows Positive Recovery Community Impact
7.0
In a 2025 study, researchers analyzed 19,685 posts from The Phoenix app, retaining 17,617 posts for analysis. They used BERT-based topic modeling and VADER sentiment analysis, identifying 10 topics across two overarching themes and a mean sentiment of 0.44 (SD 0.42), with only 2.4% highly negative. Authors conclude the app functions as a supportive recovery community and recommend larger comparative studies across platforms.
Key Points
- 1Analyzed 17,617 app posts with BERT topic modeling and VADER sentiment analysis.
- 2Identified 10 topics across two themes highlighting community support and recovery discussions.
- 3Found mean sentiment 0.44 with only 2.4% highly negative, indicating a supportive environment.
Scoring Rationale
NLP analysis of a novel recovery dataset provides actionable insights; limited by single-platform sampling and moderate topic coherence.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
