Identify data quality requirements

Disponible avec une licence Data Reviewer.

One of the challenges in implementing data quality control processes is the identification of technical data quality requirements for the organization. It is important to identify and understand the business requirements for your data before translating those into technical requirements that define good-quality data.

An effective data quality control process is based on the understanding of how data and information products are used within and outside the organization. Each organization defines quality differently and bases this definition on the intended purpose and use of the data. The following diagram illustrates a variety of sources for quality requirements that may be applicable to your organization.

Data quality elements

Data quality elements describe a certain aspect required for a dataset to be used and accurate. GIS data has different components to its quality. As defined by the International Organization for Standardization (ISO), these components include the following:

• Completeness
• Logical consistency
• Spatial accuracy
• Thematic accuracy
• Temporal quality
• Data usability

Completeness

The presence or absence of features, their attributes, and relationships in a data model.

A neighborhood with missing building footprint.

Logical consistency

A degree of adherence to preestablished rules of a data model's structure, attribution, and relationships as defined by an organization or industry. Many industries follow standards that are reflected in a geospatial data model as value domains, data formats, and topological consistency of how the data is being stored.

A highway with road surface-type gravel.

Spatial accuracy

The accuracy of the position of features in relation to Earth.

A lake feature that has been shifted.

Thematic accuracy

The accuracy of attributes within features and their appropriate relationships.

A swimming pool captured as wetland.

Temporal quality

The quality of temporal attributes and temporal relationship of features.

An outdated chart with open runway.

An updated chart with closed runway.

Data usability

Adherence of a dataset to a specific set of requirements related to a use-case.

This is used to route emergency vehicles.

This is used to map national parks.

Quality requirement documentation

A quality assurance (QA) plan is a document that identifies the quality standards that are relevant to a project and methods to achieve them. A QA plan is a living document that will change as new quality requirements are identified by the organization and also serves as an opportunity to bring together key stakeholders to build a common picture of what constitutes good-quality data and the business processes that drive those requirements.

The following are documentation techniques and standards that can be useful when identifying data quality requirements:

• ISO/TC 211 Geographic information/Geomatics—International Organization for Standardization (ISO) series of standards for geographic information to define methods, tools, and services for data management for acquiring, processing, analyzing, accessing, presenting, and transferring such data in digital form among users, systems, and locations.
• Requirements Traceability Matrix—A document created to manage and track business requirements to ensure they are met during a project implementation. This document correlates business requirements collected for the project and capabilities of a software product.

The Requirement Category column in the following table illustrates an example of collected requirements that reference some of the data quality elements outlined above. The next step after organizing and categorizing your requirements will be to correlate data quality requirements to corresponding capabilities found in ArcGIS.

Sample Requirements Traceability Matrix

ID	Requirement	Requirement number	Requirement category	Product capability
1	Ability to run queries based on number of segments edited by an individual user	F001	Functional Requirement
2	Ability to ensure the production data model is compliant with industry schema standard	D001	Data Requirement—Logical consistency
3	As geodatabase administrator, ability to restrict POST privileges to the DEFAULT version of a small set of admin users	F002	Functional Requirement
4	Ability to produce ad hoc reports indicating gaps in data for any attributes selected	F003	Functional Requirement
5	Ability to ensure that source data will be migrated into the production database and have appropriate domains and relationships	D002	Data Requirement—Logical consistency
6	Ability to ensure that source data is accurate according to the defined standards	D003	Data Requirement—Spatial accuracy
7	Ability to ensure that production data is for mobile collectors and is attribute accurate	D004	Data Requirement—Thematic accuracy
8	Ability to ensure that there is no overlap between event measures during the project period of 2010–2020	D005	Data Requirement—Temporal quality
9	Ability to hyperlink a validation error with a violated business rule and provide a description	F004	Functional Requirement
10	Ability to identify the number of cells that are not populated (NULL) for each required attribute field	D006	Data Requirement—Thematic accuracy
11	Ability to identify parcels that have no overlaying building footprint features	D007	Data Requirement—Logical consistency
12	Ability to create error reports, generate Excel files, and save them to a local drive	F005	Functional Requirement
13	Ability to validate a unique ID attribute linking a parcel to matching building footprint features	D008	Data Requirement—Logical consistency
14	Ability to confirm all features are compliant with metadata standards	D009	Data Requirement—Data completeness
15	Ability to identify existing features as an error	F006	Data Requirement—Thematic accuracy
16	Ability to indicate the location of missing features as an error	F007	Data Requirement—Data completeness