A method and system of generating a transitory sentiment community. The method comprises identifying, in accordance with a supervised trained model, within agglomerated social media content data, content associated with a subject of interest and characterized in accordance with one of a sentiment expressive usage and not a sentiment expressive usage, the subject of interest defined in accordance with at least one text character string; performing, based on an unsupervised trained model in conjunction with content associated with the sentiment expressive usage, a sentiment analysis that determines a sentiment intensity rating associated with at least a portion of the agglomerated social media content data, and generating the transitory sentiment community based at least in part on the sentiment intensity rating.