Improve Contextual Awareness using Social Network Data

This proposal seeks to leverage an innovative approach to improve health outcomes and quality of life for chronic disease sufferers by creating a platform that will connect to various social media sources pull in data containing chronic disease terms process the data using natural language processing technologies and create cohorts using both structured metadata and unstructured text data At a high level the data is sorted into categories for each chronic disease Then the cohorts are created under these top level categories Below each of the cohorts topics and themes are discovered using automatic text clustering categorization topic modeling and classification techniques for analysis By utilizing the insights gained from this analysis a greater understanding of treatment plans and self management can be gained Phase I activities include setting up the infrastructure and performing research tasks that will lay the foundation for a prototype of a cohort builder that will demonstrate the creation of at least one cohort During Phase the full architecture will be built that is capable of the broader categorization exercises Possible Phase II activities include performing sentiment analysis to gain greater insights from the data