We provide a detailed case study of text mining for knowledge discovery from social media data using a standardized terminology. We describe the steps of text mining data from the social media platform CaringBridge using the Omaha System. The text mining process included data selection, preprocessing, transformation, lexical analysis, and classification, performing at an 89% accuracy. We also describe limitations in recall and our ongoing methodological approach to increasing the effectiveness of text mining validation and lexicon augmentation. The case study is an exemplar of semantics-powered text mining. It illuminates some of the challenges of analyzing text data, such as ambiguity, idioms, and typographical errors, and presents strategies to overcome some of these challenges. It also demonstrates the value of a comprehensive, whole-person standardized terminology to support text mining.