| dc.description.abstract | This study provides a first-time systematic review into the concept of a Spatial Knowledge Infrastructure (SKI) in comparison with the well-established Spatial Data Infrastructure (SDI) concept. Generalized, an SDI is a data infrastructure where spatial data can be accessed by simple queries whereas the SKI should provide knowledge in the form of ready to go answers on questions regarding the spatial data available in the infrastructure. The systematic literature review on SKI results in four topics that were used to compare the concept with SDI: the definition, objectives, components, and architecture. Per topic, the identified differences between SKI and SDI could be classified into two perspectives: a data perspective and a user perspective. This resulted in a conceptual framework consisting of a set of parameters – data parameters and user parameters – that can be used to evaluate whether an infrastructure can be defined as SDI, SKI or something in-between. In total, 14 parameters were identified of which seven were considered to be a data parameter and seven to be a user parameter. The data parameters included the standards, producer, storage, update method, update frequency, spatial dimension, and temporal dimension. The user parameters were the level of expertise, query input, output suitability, output readiness, device readiness, analytics, and modeling.
In the second part of the study, two real-world case studies, the Kadaster Knowledge Graph and 3D Amsterdam were evaluated with the use of the set of parameters. For most of the parameters, at least one of the cases was classified as in-between SDI or SKI, or SKI-ready. With respect to the parameters update frequency and device readiness, both cases were still considered to be an SDI. Based on the findings, it can be concluded that they partially meet the criteria for being an SKI in order to fulfill present needs in the spatial domain, although it must be stressed that both cases were still under development and not fully operational. Besides, some bottlenecks already existing in SDI are not addressed. For instance, the dependency on the data quality and external data sources. The novelty of the SKI concept is reflected by the available scientific literature as the concept is predominantly discussed from a bird’s eye view. Therefore, further research is advised to parse the concept in-depth.
In conclusion, SDIs are moving towards the envisioned SKI and the proposed set of parameters can serve as guidance for SDIs to improve into a knowledge-based system that is meeting current trends in automation and modeling. | |
