Our current health applications do not adequately take into account contextual and personalized knowledge about patients.
In order to design ``Personalized Coach for Healthcare'' applications to manage chronic diseases, there is a need to create a Personalized Healthcare Knowledge Graph (PHKG) that takes into consideration a patient's health condition (personalized knowledge) and enriches that with contextualized knowledge from environmental sensors and Web of Data (e.g., symptoms and treatments for diseases). To develop PHKG, aggregating knowledge from various heterogeneous sources such as the Internet of Things (IoT) devices, clinical notes, and Electronic Medical Records (EMRs) is necessary.
In this paper, we explain the challenges of collecting, managing, analyzing, and integrating patients' health data from various sources in order to synthesize and deduce meaningful information embodying the vision of the Data, Information, Knowledge, and Wisdom (DIKW) pyramid. Furthermore, we sketch a solution that combines: 1) IoT data analytics, and 2) explicit knowledge and illustrate it using three chronic disease use cases -- asthma, obesity, and Parkinson's.
|Full Text|| |
Amelie Gyrard, Manas Gaur, Krishnaprasad Thirunarayan, Amit Sheth and Saeedeh Shekarpour. Personalized Health Knowledge Graph. 1st Workshop on Contextualized Knowledge Graph (CKG) co-located with International Semantic Web Conference (ISWC), 8-12 October 2018, Monterey, USA.