Integrity Constraints

Spatial Data Quality

Spatial data quality is most often discussed under the headings of correctness and accuracy. Correctness concerns the consistency between the data and the original source about which the data are collected and the completeness of the data itself. Accuracy has several components, including accuracy of attribute values and spatial and temporal references.

Sources of error include data collection and compilation, data processing, and data usage. The following are types of error that occur:

• Positional errors occur when the coordinates associated with a feature are wrongly recorded
• Attribute errors occur when the characteristics or qualities of the feature are being wrongly described.
• Logical inconsistencies occur in instances such as the failure of road centerlines to meet at intersections.
• Completeness, in addition to missing data, is often compromised in cases where data have been simplified. For example, when storing data about land use, the designers of the system have a choice about how fine-grained their distinction between classes of data will be (i.e., how many classes to record or how precisely the classes are demarcated), and if they choose a coarse-grained distinction (few classes, imprecise boundaries), some data may be lost.

Finally, in the final product, errors often arise when fitness for purpose is not considered. For example, the error of logical inconsistency regarding road center lines failing to meet at intersections would not be a problem for a marketing agent who just wanted to identify addresses along a road, but it would be a problem for a local government trying to map traffic flow along a road.

To be able to protect the reputation of the data provider, it is important that the user can accurately assess the quality of the data, thus minimizing the provider's exposure to risk of litigation and reducing the likelihood of product misuse. It is now becoming more common for data providers to furnish their clients with metadata (data about data) on quality, lineage, and age.

Integrity Constraints in Nonspatial Data

To preserve data quality, as part of the database design process, integrity constraints are defined. Some constraints can be specified within the database schema (or blueprint) and automatically enforced. Others have to be checked by update programs or at data entry. Integrity constraints can be subdivided into static, transition, and dynamic

...