Dr Adel Bolbol, GeoINT Consultant at Envitia, presented a paper on “Handling Uncertainty in Spatial Data” as part of the Data Quality Domain Working Group meeting. This was the final presentation by Envitia staff contributing at the OGC Technical and Planning Committee Meeting in September 2015, Nottingham.
The work presented aims to develop models, techniques and procedures for Handling Uncertainty in data capture, production, dissemination and visualisation and demonstrates how this could be managed and delivered. The research carried out looks into different stages of handling uncertainty comprising of Modelling, Capturing, Communicating and Visualising Uncertainty.
The work first develops a model that captures different aspects of data uncertainty including its dimensions (spatial, temporal, topological, thematic, documentation), sources (e.g. accuracy, ambiguity, vagueness, resolution, trustworthiness, and more) and types (collection, analysis, presentation).
The capturing uncertainty piece provides an in-depth investigation into the relationships between elements of the uncertainty model. These elements were also mapped to relevant ISO standards in order to identify links and gaps between the uncertainty model and standards meant to capture aspects of uncertainty. International standards such as MGMP, ISO 19115-19157, ASRP, WMO Core, S-57, S-100 & DGC Metadata were used to map these uncertainty elements to one another, which in turn would facilitate the exchange of this information in machine-to-machine operations.
For communicating uncertainty, the work develops a framework to quantify uncertainty based on its lineage and propagation which mainly relies on the situation/role preferences of use. The framework handles and processes uncertainty in order to communicate it in a measurable form where a decision maker could easily understand and base the critical decisions upon. The framework uses the uncertainty model elements that were defined earlier, to process the data accordingly using techniques that vary from probabilistic models to Fuzzy Sets.
As part of the visualisation of uncertainty work, the research develops a standardised method of exchanging and analysing uncertainty, as well as developing a ‘toolbox’ of visualisation techniques to represent it. Moreover, Envitia has launched an online experiment to understand the perception of different aspects of uncertainty and the use of various visualisation techniques. More information about the online experiment will be available on the Envitia’s Blog site soon.
Stakeholders from different parts of the data continuum from the Defence sector have been approached (collectors, processors, users) to capture and refine the mapping of uncertainty elements in spatial data to specifications used by their respective organisations, as well as other uncertainty aspects that do not exist in data standards. The presentation, along with others in the same DWG, emphasises the importance of understanding and appropriately handling data quality issues in order to enhance the communication of these issues; resulting in a better informed decision making mechanism.
Dr Adel Bolbol, GeoINT Consultant