Social Media Data Mining of Anti-Tobacco Campaign Messages – A Facebook Study (Preprint)

Abstract

BACKGROUND Social media platforms provide a valuable source of public health information, as one-third of US adults seek specific health information online. Many anti-tobacco campaigns recognized such trends among youth and have shifted their advertising time and effort toward digital platforms. Timely evidence is needed to inform the adaptation of anti-tobaacco campaigns to changing social media platforms. OBJECTIVE In the present study, we conducted a content analysis of major anti-tobacco campaigns on Facebook using machine learning and natural language processing methods as well as a traditional approach to investigate the factors that may influence effective anti-smoking information dissemination and user engagement. METHODS We collected 3,515 posts and 28,125 associated comments from seven large national and local anti-tobacco campaigns on Facebook between 2018 and 2021 including The Real Cost, Truth, and CDC Tobacco Free (formally known as Tips from Former Smokers), Tobacco Prevention Toolkit, Behind the Haze VA, Campaign for Tobacco-Free Kids, and Smoke-Free US. Natural language processing methods were used for content analysis including parsimonious rule-based models for sentiment analysis and topic modeling. Logistic regression models were fitted to examine the relationship between anti-smoking message framing strategies and viewer responses and engagement. RESULTS We found that large campaigns from government and non-profit organizations had more user engagements compared to local and smaller campaigns. The Facebook users are more likely to engage in negatively-framed campaign posts. Negative posts tended to receive more negative comments (OR= 1.40, 95% CI 1.20 - 1.65). Positively framed posts generated more negative comments (OR = 1.41, 95% CI 1.19 - 1.66), as well as positive comments (OR = 1.29, 95% CI 1.13 - 1.48). Our content analysis and topic modeling uncovered that most popular campaign posts tended to be informational (i.e., providing new information), where the key phrases included talking about harmful chemicals (14.3%), as well as the risk to pets (6.3%). CONCLUSIONS Facebook users tended to engage with anti-tobacco educational campaigns more that are framed negatively. The most popular campaign posts are those providing new information, with key phrases and topics discussing harmful chemicals and risks of second-hand smoke for pets. Educational campaign designers can utilize such insights to increase the reach of anti-smoking campaigns and promote behavioral changes.