ABSTRACT
Although data quality has long been accepted as an essential component of geospatial data, as indicated by the history of spatial data quality research in the work of Devillers et al. (2010), proper use of the acquired geospatial data remains a significant challenge for current geographic information system (GIS) users. The ability of GISs to integrate data of various qualities can lead to errors in the final results (Lanter and Veregin, 1992). Traditional GISs rely heavily on experts to understand the data they use and ensure that every decision is made correctly. As the sharing of geospatial data has become easier with the emergence of geoportals (Maguire and Longley, 2005), the implicit or explicit discrepancy and heterogeneity of data quality between data sets acquired from various georesources have become major obstacles to data interoperability. Given the lack of a comprehensive framework for modelling, distributing, analyzing, and visualizing the quality of heterogeneous geospatial data, GIS users are forced to deal with data of ambiguous or even unknown quality. Thus, an unpredictable level of risk that users may never even notice is inevitably hidden in the final decisions. Given that GIS functions are often naively used to process, analyze, and derive new information, we argue that the awareness of data quality must be considered during the design of GIS functions. Otherwise, users will be working in an extremely risky application environment and cannot take full advantage of the resource sharing brought by spatial data infrastructure (SDI).
