TopBraid Teamworks Framework
Change management for Knowledge Graphs
The Teamworks Framework is part of the core technology infrastructure underlying TopBraid EDG. The framework has two distinct features:
Change Management with complete audit trail and support for workflows with associated working copies or sand-boxed versions
Model-specific Applications – ability to associate different editing applications, reports, export, imports and other features with different graph types aka Asset Collection types.
Change Management with Teamworks
Teamworks manages each asset collection using two kinds of graphs:
- The master graph that contains the actual data
- The TCH (or team) graph containing the audit trail, working copies, user permissions and comments.
As users make changes, audit trail of changes is automatically written into a TCH graph. An audit trail includes information about the change, who made it and when. This information is available in the EDG user interface within the Change History panel. It can also be queried in SPARQL or GraphQL. Master graph is also changed accordingly, unless a user is making changes in a context of a workflow.
Each workflow in TopBraid EDG has an associated working copy. A working copy is a Teamworks feature for managing a set of changes in a sandbox, without writing them to the master graph. All changes are stored only in the TCH graphs making the working copy a virtual view that is computed on demand. Although a working copy is only a computed virtual view it has identity and can be interacted with by users and APIs as if it was physically stored.
When a workflow completes successfully all changes it carried are written into the master graph aka production copy. The audit trail of such committed changes continues to be available after the commit.
Teamworks Support for Graph-specific Editors and Operations
The best way to understand this feature is by examining TopBraid EDG. It is shipped with many predefined graph or asset collection types. Each asset collection type has a declared editing application as well as declared reporting, import/export and other features. For example:
- Taxonomies – these asset collections use an editing application that is optimized for using SKOS. They have unique import, export and report options such is import from MultiTes and hierarchy export.
- Ontologies – these asset collections have editing UI designed for working with schemas – classes, properties, shapes and rules
- Data Asset Collections – these asset collections are using a tabular editing application targeted to cataloging data assets. Specialized import options enable ingesting metadata from data sources.
TopBraid EDG customers can define their own custom types of asset collections. Internally, they are called “EDG project types”. For example, a new asset collection type may be focused on capturing information about a decomposition of engineering assets or about business process flows. Using this feature requires licensing of the Data Graphs module.
When a new type of collection is created, it will show up as a new link in the EDG navigation bar – on the same level as asset collection types shipped with EDG such as Taxonomies, Ontologies and Data Assets. With the new type established, users can define an edit application specific to the new type and specify other type-specific options such as imports or data transformations. All these definitions are made declaratively as part of the knowledge graph.
Ready to get started?
Get in touch today to learn how to improve semantic data governance for your enterprise.