Configuration
Configuration
Breadcrumbs

Metamodel and metadata templates

The metamodel in Anjana Data constitutes the conceptual foundation on which the organization's various data and information assets are governed. Its configuration allows explicitly defining:

  • The entities that will be managed (e.g. datasets, reports, AI models, business terms).

  • The relationships between those entities (e.g. trained with, composed of, used).

The correct definition of the metamodel is fundamental to:

  • Represent the reality of the data in the organization in a clear and structured way.

  • Establish a common framework that supports the governance and asset management processes.

  • Ensure consistency in navigation, search, lineage and auditing within the Data Portal.

This section details the steps and considerations needed to configure the different components of the metamodel:

  • Entity subtypes: definition of the types of objects to be governed (e.g. datasets, reports, AI models, business terms).

  • Relationship subtypes: definition of the links between entities that reflect dependencies, hierarchies or associations (e.g. trained with, composed of, used).

Together with the metamodel, the metadata templates allow parametrizing the way in which data assets and their relationships are described and documented on the platform. These templates include the creation and adjustment of:

  • Menus and sections to structure the information.

  • Attributes of different types (text, numeric, taxonomies, lists, etc.).

  • Validations and versioning rules that ensure quality, consistency and impact control in data management.

Important notes

  • The metamodel must remain relatively stable over time, since it reflects the organization's semantics and business logic.

  • Changes to the metamodel structure can directly impact:

    • The history of governed assets.

    • The lineage of the data.

    • The configured workflows.

    • The permissions configuration.

    • The user experience in the Data Portal.

  • Anjana Data includes a native metamodel, which provides logic and support for the platform's own functionalities.

Anjana Data native metamodel

Anjana Data includes a native metamodel that defines a set of entities and relationships with an internal logic within the platform. This metamodel provides the minimum foundation necessary for the data portal to work coherently from the start and for administrators to be able to extend and adapt it according to the organization's governance needs.

Native entities

The entities natively governed in Anjana Data are:

  • DATASET: A set of structured, semi-structured or unstructured data - persistent or transient - that can represent database tables, files in data lakes, documents, images, audio or video.

  • DATASET_FIELD: A field or column of a structured dataset, with its own metadata, that allows governing the data at the individual attribute level.

  • DSA (Data Sharing Agreement): data sharing agreement or contract.

  • PROCESS: business or technical process that manages data.

  • INSTANCE: specific execution or instance of a process.

  • SOLUTION: solution or application that groups and manages instances and processes.

Native relationships between native entities

The native metamodel also includes the following predefined relationships:

  • STRUCTURE: between a dataset (source) and its dataset fields (target).

  • DSA_CONTENT: between a DSA (source) and the entities it contains (target).

  • ADHERENCE: between a DSA (source) and a User (target).

  • INSTANCE_PROCESS: membership of an instance (target) to its process (source).

  • INSTANCE_DATASET_IN: between an instance (target) and the datasets it reads from (source).

  • INSTANCE_DATASET_OUT: between an instance (source) and the datasets it writes to (target).

  • SOLUTION_RELATED_INSTANCE: between a solution (source) and its related instances (target).

  • SOLUTION_OWNED_INSTANCE: membership of an instance (target) to its owning solution (source).


Important note:

  • These entities and relationships are part of the native core of Anjana Data and provide the logic needed for basic platform functionalities, such as search, lineage and adherence control.

  • The administrator can extend and customize the metamodel by adding new entities, relationships and rules, always respecting consistency with this base metamodel.

Order for configuring the metamodel and the metadata templates

Once the native metamodel of Anjana Data (predefined entities and relationships) is understood, it is important to follow a logical order when configuring the extended metamodel and the metadata templates.

The recommended process is as follows:

  1. Metamodel - Entities and relationships: definition of the object subtypes (entities and relationships) that will make up the metamodel.

  2. Template menus: creation of the main top-level blocks that organize the information in dynamic forms.

  3. Template sections: configuration of the subsections within each menu, which allow grouping related attributes.

  4. Metadata attributes: definition of the attributes that describe the entities and relationships, reusable across different templates.

  5. Reference values: parametrization of the values available for selection-type attributes (lists, taxonomies, icons, etc.).

  6. Assigning attributes to templates: inclusion of the attributes in the corresponding sections of each template.

  7. Validations on templates: definition of rules to ensure consistency and restrictions of the attributes.

  8. Relationships between reference values: establishment of dependencies between reference attribute values or hierarchical structures in taxonomies.

  9. Versioning rules: configuration of the changes that generate a new version of the governed objects.

  10. Workflow-trigger exception rules: definition of exceptions that prevent the triggering of approval workflows on specific modifications or versioning.

This ensures an orderly, coherent and sustainable configuration of the metamodel and the metadata templates.