Download Full Text (265 KB)



Social media platforms facilitate user interaction and impact decision making. Users prefer to use hashtags while sharing posts. Knowing the sentiment towards diabetes, bloodpressure, and obesity is fundamental to understanding the impact of these information on patients and their families. The study seeks to determine the relevance of health-related hashtags on Twitter and analyze sentiments about diabetes, obesity, blood pressure.


Tweets were retrieved using synonyms for “diabetes”, “hypertension” and “obesity”. The extended knowledge discovery in data mining (KDDM) model guided our research with research objectives defined in the ‘research problem understanding’ phase. The ‘information seeking’ from Uses and Gratifications Theory (UGT) determined the success and text mining assessment criteria. Text pre-processing was done using tokenization, stop word removal, and stemming. The research objectives, text mining goals, and success criteria were answered using ‘Uses and Gratifications Theory’ (UGT).


Total 6749 tweets were extracted using RStudio. 36.41% were about blood pressure, 0.25%- diabetes, 24.43% -obesity and 6.99% -combination of two or more terms. Additional topics such as cholesterol, chia seeds, postpartum, diet, exercise were identified. Upcoming conferences like ‘#ipna’, ‘#review’, ‘#APCH2019’, ‘#cardiotwitter’ were identified. Increased user engagement – about managing blood pressure, diabetes, obesity across different age groups, as well as the consequences of increased cardio exercise for obese and diabetic users were encouraging. Tweets about advertisements specific to clothing for oversized individuals-initiated conversation among users about monitoring self-health.


Sentiment analysis can thus increase our understanding about user engagement on such platforms and potentially help improve managing public health strategically.

Publication Date



Management Information Systems

Is Part Of

VCU Graduate Research Posters

Relevance of Health-Related Hashtags on Twitter: A Text Mining Approach