Basic terms

Data model

Full description. An informational structure of data, which allows you to specify what data in what form, and in what location will be stored. The model consists of objects: entities and lookup entities. The data model as an informational structure also includes data sources, units and enumerations. When loading data from other systems into Unidata, all data is arranged in a data structure.

Short description. A description of the data structure, containing entities/lookup entities, as well as units, enumerations and data sources.

Entity

A data model object that contains data in the form of records with a set of attributes. Data in the entity may change or be supplemented over time. The entity allows you to create complex attribute structures and supports setting relations with other entities/lookup entities books.

Examples: “Delivery zones”, “Office equipment”, “Customers”.

Lookup entity

A data model object that contains data in the form of records with a set of attributes. The data in the lookup entity is rarely changed. The lookup entity does not support setting relations.

Examples: glossary of terms, manufacturing or product standards.

Unit of measurement

A description of the quantities to be measured and their parameters. Any unit of measurement contains a basic unit to which all others are reduced.

Examples: volume, currency, mass. The “Mass” basic unit could be a kilogram. All attribute values in grams, centners, and tons will be converted to kilograms when searching for records.

Data source

Third-party information systems from which data is sent to the system. Each data source has its own level of trust (weight), which allows you to resolve conflicts in the data.

Example: A source system with data from CRM has a weight of 80; a system data source has a weight of 100.

Enumeration

A set of possible values of the attribute.

Example: The “Order status” attribute allows you to choose one of the values: wait for payment, in process, delivered, etc.

Relation

The relationship between records in different entities/lookup entities. Relations can be of different types.

Example: The “Producer” attribute of the “Goods” entity has a relation with the “Country name” attribute of the “Countries” lookup entity.

Record

Description of the object and its characteristics. Characteristics of an object are described by attributes.

A record is data placed in an informational structure (in an entity/lookup entity).

Example: In the “Customers”entity, the records will contain all information about customers, their details, contacts, etc.

Etalon record

A record that can be considered up-to-date and free of errors, inconsistencies, etc. Typically, an etalon record can be obtained by checking it with data quality rules, combining it with potential duplicates, obtaining data from reliable source systems, etc.

The etalon record may also be called the Golden record.

Attribute

A characteristic of an object that has a name and a value. Attributes may have different types, depending on their purpose.

Example: The “Producer” attribute with the “Japan” value.

Validity period

A period of time during which a version of a Record is valid. If there is more than one period, the record is valid for the period that comes up on the current date.

Example: The version of the record with the promotional price of a product is only valid in the period from 01/26/2010 to 03/31/2010.

Data quality

The level of suitability of the data for business: a measure of how much data is incorrect, what is the level of discrepancies, inconsistencies in the data, etc.

Quality rule

A rule to which the data must match. A quality rule contains a data processing function and an operation mode. If the data does not match a quality rule when created or modified, a data quality error is created.

Examples: Serial number checking, auto-removing of extra spaces.

Data quality error

A message telling that the data does not match a quality rule. The error may contain an error message, error level, etc.

Example: Date format is incorrect or the record attribute is empty.

Validation

Checking and validating data against pre-defined quality requirements. Validation is used to identify meaningless, uninformative, incorrect, or erroneous data.

A quality rule in validation mode creates quality errors.

Example: validating a phone number format.

Data enrichment

The modification of data according to certain rules during the operation of quality rules. Enrichment is intended to modify the data so that it is completed unified, or supplemented with new information.

Example: For the “Customer passport” entity, the value of the “Customer” attribute is formed from the values of the “Customer name”, “Shipping address” and “Contacts” attributes of the “Customers” entity.

Data processing function

An action, or sequence of actions, that is performed on data. Functions are used in quality rules and allow you to either check incoming values or modify them.

Examples: sum two numbers, check if an attribute is completed.

Security label

A security setting that allows you to restrict user (user role) access to certain attributes of entity/lookup entity.

Example: A security label hides attributes with personal customer data from a particular user role.

Workflow

A process description intended to approve any changes according to a preconfigured algorithm (based on BPMN notation). This is a separate component of the platform that allows you to perform the required tasks in the process (e.g., record changes approval).

Example: A workflow for three-step record approval for the “Raw material purchasing” entity.

Example of workflow

A copy of the workflow that is created when changes are initiated. For example, if a workflow is configured for the “ Producers” entity, when a steward attempts to edit a record of that entity, a workflow example will be created, which will result in the approval of record changes.

Example: Workflow for three-step approval of the “Sand purchasing” record for the “Raw materials purchasing” entity.

Task

A request to make changes to the data. Represents a workflow step and is created automatically (for example, when publishing a record draft). Starting a workflow creates as many tasks as there are steps in the process.

Example: A task to approve record by the data stewards team lead.

Approval

A process to confirm or reject changes made to the data. The approval is performed in the task.