OpenStreetMap data assessment to support health services in the Global South, The cases of Wajir and Nairobi counties, KE

Abstract

The present research was an attempt to evaluate OpenStreetMap’s capacity to support health services in the Global South. A user-oriented workflow was built so that the assessment of data is not generic anymore as oftentimes in the literature. Instead, a fitness-of-use framework was employed for a more precise assessment. This framework was applied in two use cases, the Counties of Wajir and Nairobi in Kenya. The questions that had to be answered concerned first of all the needs of the health workers and the capacity of the OSM vocabulary in the areas of the use cases to satisfy them. Secondly, the research tried to discover if semantic web technologies can enhance that capacity by unifying a fragmented vocabulary and even help build an ontology to promote it. Then the possibility of a proxy for the quality of OSM data through their correlation with humanitarian mapping projects was investigated. Finally, the basic setup of a web application that supports health workers and individuals in the Global South using OSM data was sought after. The use cases chosen were quite different in socioeconomic characteristics. For Nairobi County, it was discovered that the OSM vocabulary has the capacity to reflect the needs of health workers while Wajir is far poorer. The use of triplified OSM data and the SPARQL queries allowed sophisticated searches for objects and attributes that then were easily translated into ontologies. Through those ontologies, it became possible to query for information that was not tagged in OSM. Moreover, it was shown that the existence of completed humanitarian mapping projects in an area has a positive linear correlation with the completeness of health facility objects and some of their attributes. In conclusion, the present research provided a basic setup of a web application that supports health services with the use of OSM data.