A knowledge graph for eldercare: Constructing a domain entity graph with guidelines

Abstract

The sheer size of aging population and the complexity of individual eldercare needs have put great social and economic burdens on various societies worldwide. Particularly, the daunting demands for caring for elders are unmet by the overall supply of caring capacities. There are at least two aspects of this imbalance: first, the number of high quality caring professionals with nursing backgrounds have been insufficient and this trend will continue in the near future; second, most of the current professional and semi-professional caregivers have limited education that in turn limited their caring capacities. To tackle these issues with the rapid development of data science and growing awareness of nursing informatics, we are working on building a knowledge graph for eldercare. In this paper, we propose the concrete first step towards this goal: constructing an eldercare entity graph based on international guidelines, and this is first of its kind to our best knowledge. Protégé is used to construct the entity graph that consists of eight types of entities and four types of relationships. In this graph, a standard eldercare procedure is represented as a path between an entity denoting an eldercare service classification and an entity denoting an eldercare activity. About 230 procedures could be derived according to the guidelines.