Objectives: The goals of this study were to examine the feasibility of using ontology-based text mining with CaringBridge social media journal entries in order to understand journal content from a whole-person perspective. Specific aims were to describe Omaha System problem concept frequencies in the journal entries over a four-step process overall, and relative to Omaha System Domains; and to examine the four step method including the use of standardized terms and related words. Design: Ontology-based retrospective observational feasibility study using text mining methods. Sample: A corpus of social media text consisting of 13,757,900 CaringBridge journal entries from June 2006 to June 2016. Measures: The Omaha System terms, including problems and signs/symptoms, were used as the foundational lexicon for this study. Development of an extended lexicon with related words for each problem concept expanded the semantics-powered data analytics approach to reflect consumer word choices. Results: All Omaha System problem concepts were identified in the journal entries, with consistent representation across domains. The approach was most successful when common words were used to represent clinical terms. Preliminary validation of journal examples showed appropriate representation of the problem concepts. Conclusions: This is the first study to evaluate the feasibility of using an interface terminology and ontology (the Omaha System) as a text mining information model. Further research is needed to systematically validate these findings, refine the process as needed to advance the study of CaringBridge content, and extend the use of this method to other consumer-generated journal entries and terminologies.
Bibliographical noteFunding Information:
Financial support for this research was provided by CaringBridge, a non-profit located in Eagan, Minnesota.
© 2018 Faculty of Health and Social Sciences of University of South Bohemia in České Budějovice
Copyright 2020 Elsevier B.V., All rights reserved.
- Omaha System
- Social media
- Text mining