Understanding conversations on alcohol across diverse Reddit communities: a computational analysis

Mansi Shah, Yara Acaf, Michael Mackert

Abstract: This study examines how diverse population segments within a U.S. state discuss alcohol-related issues on Reddit. Using computational methods, data were collected from three subreddits via the Reddit API. Textual and sentiment analyses were conducted on user comments, followed by machine learning techniques to identify frequently used words and topics. The results highlight distinct patterns across the subreddits. The findings underscore the cultural diversity within the same state and a sobriety-focused community. The insights have significant implications for tailoring statewide health communication campaigns. By addressing unique regional and cultural themes, campaigns can achieve greater relevance and effectiveness in reaching target audiences. This research highlights the value of leveraging social media data for nuanced, localized health communication strategies.

External IDs:dblp:journals/jocss/ShahAM25