Content and Context Based Analysis of University Social Media Engagement

Social media has become an essential tool for institutions to create interactions with their users. Creating successful and relevant material social content is challenging as struggle comprehend what drives user engagement. Using analytics on user-generated determine the relation between post features engagement becomes a method overcoming this challenge. In study, measured based quantitative approach number of likes comments. We applied machine learning algorithm extract contextual (time post) (caption image classification) from university Instagram posts, influence statistically modeled. LASSO regression applied, which effective strategy automated variable selection. The results indicate that context around 13:00 Sunday, during month June, containing words "Gedung" "Foto," well classifications "Sight" "Sadness," will receive more