Developing a quality assurance plan allows you to define the acceptance criteria for your data quality as well as the necessary tests and processes required to ensure quality data. While you may follow your own validation process, a typical quality control life cycle has three stages, all of which are supported by ArcGIS Data Reviewer.
Validation
Data Reviewer supports two levels of validation: Detect, which is an automated validation, and Review, which is a more in-depth visual or manual review of the data.
Detect—Within Data Reviewer there are more than 40 different types of automated validation checks that can be configured for your data (see the Data Reviewer Checks poster). These checks include:
- Validating attributes to ensure the values match a pattern, such as an identifier must have two letters followed by four digits.
- Validating the integrity of features to ensure collection rules are met, such as the minimum length of a line feature.
- Validating the spatial integrity of your data, such as finding gaps or overlaps between features.
Review—While many automated validation techniques exist, often a more in-depth visual or manual review of the data is required; for example, to find missing features. Data Reviewer supports visual or manual review in several ways including:
- Performing a visual review using symbology and labeling to identify attribution that is missing or incorrect.
- Using a polygon grid to divide your data into smaller sections for detailed evaluation and tracking sections you have reviewed.
- Using a sample set to randomly select a representative set of features. For more information about different sampling methods, see Calculation methods for sampling.
Correction
Data Reviewer aids in the process of correcting data by providing an easy-to-use interface — the Reviewer table — to store the anomalies identified during the detect/review stage. Clicking on a record in the table zooms or pans you to the anomaly. The Reviewer table allows you to record when and who resolved the anomaly, at which point you are ready for the verification stage.
Verification
The final stage in the quality control life cycle is to verify the anomalies that have been resolved according to your business rules. The same validation rules used to find the anomalies can be re-executed to ensure that they have been resolved and no new anomalies have been introduced. The Reviewer table can then be used to navigate to the anomaly records and track their verification status.