Key Terms & Abbreviations
In the following section, we will define the key data quality terms that we will refer to throughout the Data Quality Framework.
Attribute: Single unit of data that in a certain context is considered indivisible. See Data element.
Business Area: A specific segment of a company’s operations or activities.
Business Data Owner: The individual responsible for data governance and management within a business domain.
Business process owner: Person or organization who is accountable for a given business process. In some cases, the Business Process Owner can be the same person / organization as the Data Owner. For further details on the role, please refer to the DQF Resource Center.
Business user: Person within a business organization who uses data in his/her daily work, such as analyst and operational staff, to inform decision-making and drive business outcomes.
Criticality Level (Low, medium, high, critical): The degree of importance or urgency assigned to an issue or data set.
Data consumer: Data consumers are those in an organization who use data. This is usually everyone in the organization to a greater or lesser extent. For further details on the role, please refer to the DQF Resource Center.
Data domain: A data domain is a logical grouping of data. It can be by data object (E.g. customer or employee) or by process area (for example, commercial data or HR data).
Data element: A data element is a unit of data that has a clear meaning and precise characteristics. It can refer to a column in a table or a field in a file (E.g. name, age, profession).
Data engineer: For further details on the role, please refer to the DQF Resource Center.
Data governance team: The data governance team includes roles like ADO, BDO, Data owner and Data steward. For further details on these roles, please refer to the DQF Resource Center.
Data owner: Person or organization who is accountable for a specific set of data. For further details on the role, please refer to the DQF Resource Center.
Data producer: As opposed to Data consumer, Data producer is the term for people in an organization who create data in systems.
Data profiling: Data profiling assesses a set of data and provides information on the values, the length of strings, the level of completeness, and the distribution patterns of each column.
Data quality analyst: Person whose role includes key activities in data quality management, including data profiling, developing quality metrics, conducting audits, and using monitoring tools to track and resolve data issues. For further details on the role, please refer to the DQF Resource Center.
Data quality control: A data quality control is logic that is applied to a data element or a Dataset, to determine whether the targeted data is correct or incorrect.
Data quality dimension: Data quality dimensions are a way of organizing the various data quality rules into themes. For example, data quality rules that assess whether data is missing or not, would be mapped to the “Completeness” data quality dimension. For further details on the list of data quality dimensions used in the enterprise, please refer to the “Controls” section of the Data Quality Framework.
Data quality incident: A data quality incident is a group of similar data quality issues that share the same properties.
Data quality issue: A data quality issue is a single data quality check result that was not accepted by a data quality rule. The percentage of data quality issues within the total number of data quality checks performed is also used to measure the overall quality of data by calculating the data quality KPIs.
Data quality level: A classification system indicating the quality standards of data (E.g. ‘C1 - Gold’).
Data quality monitoring: The process of reviewing the current status of data quality in an organization, using reports that aggregate the results of rule checks.
Data quality prioritizing: Prioritizing data quality refers to the process of identifying and organizing data quality issues in order of importance or urgency based on business impact.
Data quality remediation: The process of identifying and correcting errors and issues related to the reliability of data. It involves executing immediate measures to resolve the issue and identifying the need to implement more sustainable mechanisms.
Data quality rules: For the purpose of the Data Quality Framework, data quality rules are considered to be the same as data quality controls.
Data quality self-assessment: The process of evaluating a team’s / organization’s practices in place with the objective to self-reflect on strengths, weaknesses, achievements, and areas for improvement.
Data quality sustainability: The process of maintaining a good level of quality for Data over the long term by tackling root causes and continuous improvement.
Data quality team: Refers to a team that includes technical and functional roles that are in charge of the activities such as data profiling, data monitoring, data remediation. Depending on the context they might be part of R&D teams, and made data quality analysts, data engineers, developers, etc.
Dataset: A dataset is a logically meaningful grouping of data organized in a defined format. Datasets are typically organized in a structured format such as rows and columns, like a table or a file (for example active customers’ dataset).
Data steward: Person with key data management responsibility on a specific set of data. His/her role includes ensuring metadata accuracy and technical data lineage, executing data quality requirements, applying standards and policies, and supporting data owners with necessary information. For further details on the role, please refer to the DQF Resource Center.
External stakeholder: Customers, partners who may rely on the enterprise’ data and who contribute by providing feedback and helping to define data quality expectations based on their needs and preferences.
Failed row: Failed rows are rows of data that have been run through a data quality rule and deemed incorrect – E.g. A record that is supposed to contain a value but is actually found to be empty. The inverse, would be known as a past record.
Metadata: Data defining and describing other data.
Score: A numerical value representing the quality or performance of data.
Threshold: The set limit or value at which an action or decision is triggered.