EDG 8.0 Release

TopQuadrant is excited to announce TopBraid EDG 8.0. With a focus on simplifying knowledge graph (KG) creation and curation using vector databases, and enhancing scalability and stability by integrating with graph databases, EDG 8.0 marks a major milestone in our commitment to empowering organizations with advanced data management capabilities. 

Using LLM-tech for easier knowledge graph curation

To simplify the KG curation process, EDG 8.0 contains a built-in vector database, making it easier for users to align their ontologies and taxonomies and enabling entity resolution. This LLM-tech also enables the automated tagging or labelling of unstructured data. No more training required – this one shot approach allows users to automatically assign classes or entities from their controlled vocabularies to documents based on the underlying semantics of the content. This will allow users to integrate unstructured data from their CMS platforms, or elsewhere, into their KGs more easily. The built-in vector database will also lay the groundwork for any GenAI applications that you may want to build at your enterprise, like Retrieval-Augmented Generation (RAG). Documentation can be found here. 

Integration with graph databases

For enhanced scalability and stability, EDG 8.0 allows for out-of-the-box live integration with any graph database. Users can integrate EDG with any graph database with a SPARQL endpoint, like GraphDB or Neptune, using a user-friendly wizard. Users can read and write both ontologies and instance data stored in the graph database from their EDG interface, without bogging EDG down with all the additional data. This capability also contains ‘complete mode’ which will allow users to cache an entire graph from an external database in EDG for data validation analytics. This allows users to keep the bulk of their instance data in a graph database, taking advantage of what those products were built to do, while keeping their ontologies, taxonomies, and metadata in EDG, leveraging what EDG was built to do – helping users curate and align their KGs and controlled vocabularies. This capability also allows for multiple remote triple store integrations, allowing users to maintain well-governed data across multiple graph databases and applications, while incorporating third party data available via SPARQL endpoints (like Wikidata). Documentation can be found here.

Medical Term Management Accelerator

We have also built two accelerators – packaged solutions for use cases we have expertise in. The first is Medical Term Management (MTM). MTM is focused on helping life sciences companies maintain their internal reference data sets and ensuring that they are aligned with third party controlled vocabularies like SNOMED, MedDRA, ICD, RxNorm and many more. Maintaining changing vocabularies within and across an enterprise is hard enough, but ensuring that they are also aligned with external vocabularies that are also changing, is incredibly time consuming. MTM allows for the automated detection of changes in these external vocabularies and the change management capabilities to ensure consistent alignment. Reach out to CustomerSuccess@TopQuadrant.com if you’d like to learn more. 

Data Policy Enforcement Accelerator

Additionally, we have built an accelerator focused on policy compliance, the Policy Enforcement Accelerator. Many users rely on EDG to manage their data governance policies and how they correspond with different data assets (databases, tables, database columns). Rather than trying to align individual data assets with the policies that should govern them, it is better to map these data assets to a business glossary, and then map enterprise policies to the business terms in that glossary. That way the other database columns, for example, that map into the same business term all share the single definition of the applicable policies. This accelerator sets customers up for data governance success and more efficient policy compliance. Reach out to CustomerSuccess@TopQuadrant.com if you’d like to learn more. 

Miscellaneous improvements:

• Import classonomy from file: This feature is available from the global “New” button and is designed to handle large OWL ontologies that contain thousands of classes but no instances. This is a pattern that is often used in biopharmaceutical ontologies such as those from BioPortal. The classonomy importer converts OWL classes into SKOS taxonomy instances. The OWL class axioms will be preserved during the conversion, in case users still want to query them.
• Problems and suggestions panel now has a tabular display: This allows for easier filtering and sorting of the problems and suggestions.
• 3D graph panel: The 3D Graph panel can be used to browse and explore assets and their relationships in a three dimensional space. This can sometimes be helpful to better understand the general layout of a graph or the neighborhood of specific nodes.
• Support for programming multi-page wizards: This enables experienced EDG users to perform complex data operations with ease, enhancing overall efficiency and effectiveness.
• Support for Single Sign-on with OpenID Connect: This is part of a redesigned authentication system. Other available authentication methods, including SAML2, have been improved as well.

Numerous bug fixes and usability improvements:

Please review the change log for complete details.