Tasks for different user groups

All functions of the Unidata platform are divided between three main user groups. Each group performs its own tasks.

Data steward

Main task: data processing (i.e. individual entity/lookup entity records). For example, the “Real estate developers” entity will contain entries for each of the developers. Attributes in the record: legal entity type; name; phone number; legal address, etc.

Simple data processing tasks:

  • Enter attribute values, both for the current moment and for a specific record state in the past or future (for a different time interval);
  • Edit existing records;
  • Delete records.

The platform allows you to create and configure workflow processes that let you organize an approval algorithm for any changes made to records through tasks.

Data stewards have access to the following functions depending on the their roles and access rights sets:

  • Vieng records;
  • Edit and create records. (Includes the workflow processes reliant algorithms);
  • Coordinate record changes made by other users.

Data steward may also be restricted from viewing individual entities, lookup entities, or records in some of them.

Records can be edited for several reasons:

  1. Processing records that are partially or completely duplicated. Rules for potential duplicate search are configured separately. Records with potential duplicates get marked after the rules are applied. The steward's task is to manually compare combine these records if they match.
  2. Processing records with quality errors. Data quality rules are configured separately. Records with quality errors get marked after the rules are applied. The steward's task is to open a record, access the attributes that have errors, and correct said errors. E.g.: the phone number may be incorrect, or the last name may contain numbers.
  3. Creating, editing, or deleting records can be related to internal business tasks, such as expanding the database.

Data steward can perform batch actions to change records (e.g.: mass edits), and import/export data from Excel.

Data administrator

The main task is creating and configuring an informational data model and rules for detecting incorrect data.

A data model with entities/lookup entities is created to describe the required subject area. Each entity/lookup entity requires a set of attributes, relations, etc.

To systematize the entities/lookup entities records, create and configured the record classifiers. Attributes with additional record information can be created within individual classifier nodes. For example, a record in the “Spare Parts” entity can have 3 basic attributes, but if you you assign a “Gearbox” node to it, it will add several attributes that describe properties specific only to gearboxes.

The data model also allows the manager to set up quality rules that require basic or manually configured data processing functions. Duplicate search rules can be configured for entity/lookup entity attributes. Among other things, the data administrator can configure enumerations, measurement units, and data sources; import/export data models, including entities, lookup entities, classifiers, roles, accounts, etc. The data manager role can be partially or fully integrated with the platform administrator role.

System administrator

The main task is to maintain the system, configure the security model (see section Security model description), configure workflow processes, and audit user actions.

The platform administrator manages operations for importing, exporting, deleting, reindexing data, matching duplicate clusters, etc. The platform administrator's tasks may also include troubleshooting, data backup, and changing the basic configuration, such as changing the storage period for user activity data.

The platform administrator role can be partially or fully integrated with the data manager role. The platform administrator role can also be integrated with system administrator or engineer roles.