Some early experiences using the new Relationship Explorer feature in IBM Knowledge Catalog

Pat O'Sullivan
6 min readMay 3, 2024

--

Photo by Look Up Look Down Photography on Unsplash

This article has been cowritten with Susan Cotter and Michael Loughran

IBM Knowledge Catalog is a comprehensive Data Governance tool with an extensive set of object types to represent the specific features of Business and Technical metadata

We are part of a team in IBM development responsible for the creation and maintenance of the IBM Knowledge Accelerators , an extensive set of industry-specific business terms and other governance artifacts that provide out-of-the-box vocabularies for Banking, Insurance, Healthcare and Energy & Utilities within the IBM Knowledge Catalog tool. The Knowledge accelerators consist of a range of different Business Artifacts including Business Terms, Categories, Tags, Policies, Governance Rules , Reference Data and Data Classes.

As the breath of topics covered by each Knowledge Accelerator is extensive, and given the variety of metadata covered by each object type, seamless and intuitive navigation is essential

Enter the Relationship Explorer

In IBM Knowledge Catalog there is a new feature called the Relationship Explorer. This is probably one of the most exciting additions to the overall user experience of IBM Knowledge Catalog. Relationship Explorer gives users a highly intuitive, visual and graphical approach to viewing how the various objects in IBM Knowledge Catalog are inter-related. Relationship Explorer allows a user to build up a view of the neighbourhood of relationships for any given artifact.

How does it help our development activities

As part of the overall IBM Knowledge Catalog development organization we have had access to early versions of Relationship Explorer. From the very moment we started to use this feature, it quickly became a critical part of our tool kit in building and maintaining extensive industry content. The rest of this blog outlines some of the ways where the Relationship Explorer radically enhances the experience of ourselves and our clients when working with the Knowledge Accelerators.

Predefined Business scopes

Our Knowledge Accelerators typically consist of 1000’s of Business Terms and other governance artifacts, describing information in a specific industry. To get customers started, we provide Business Scope categories that group that information into smaller sets focused on specific business topics

Before Relationship Explorer, these scopes could only be visualized as a flat list of business terms associated with the Category representing the particular scope.

Now, by simply clicking on the “Explore Relationships” button that scope is transformed into a rich visual network of inter-related business terms, as shown below with the Capitation Agreement scope which is part of our Knowledge Accelerator for Healthcare.

Selecting from the different types of Business Term Relationships

Knowledge Accelerators provide a highly integrated vocabulary describing the business information and relationships. So certain commonly used Business Terms have a wide range of relationships to other Business Terms and governance artifacts. While it is of course possible to view these related Business Terms from the main Business Term details page, it is also possible to navigate through these different types of relationships from the Relationship Explorer. A user can visually navigate relationships and select which artifact they want to add to their Relationship Explorer canvas. In this way, it allows them to build up a picture of the information network that would otherwise need to be acquired from multiple details windows

In the example shown below of the term Individual from our Knowledge Accelerator for Financial Services, you can also view the different types of Term to Term relationships from the menu on the left hand side. However, it is also possible to start to build up a view of the set of different relationships to that term that are of particular interest to the user at a particular point in time. In the example below, the user had used Relationship Explorer to investigate some connections from Individual to an aspect of GDPR, and to Suspicious Activity reporting

Viewing Type Hierarchies

Knowledge Accelerators use the “Is type of” relationship to define the classification hierarchy of major business concepts such as the decomposition of product types, parties, business activities, etc.

By zoning in on the “Is type of” relationships it is possible to establish the relevance and scope of requirements to areas of their business. It is also possible to see a complete picture of the hierarchy and branch out into other relationship types as required. This allows users to build a picture of inherited characteristics on one screen that would be difficult to ascertain across multiple details windows.

In the example below we see a subset of the Involved Party type hierarchy from our Knowledge Accelerator for Financial Services.

Understanding the scope of Regulations and Standards.

A critical use case for our Knowledge Accelerators is to understand the potential scope of various industry regulations and standards. The Knowledge Accelerators include an optional import of large numbers of business terms relating to various regulations. Having a way to quickly and visually see the integration of these separate taxonomies of terms with the Knowledge Accelerator Core Vocabulary is critical.

In the image below, we have taken a small slice of our GDPR taxonomy and used the Relationship Explorer to see the integration from Data Subject in GDPR to Individual in our core vocabulary. We would represent such integrations to Regulatory Taxonomies via custom relationships (the relationship Assigned to in the example below), which are also supported by the Relationship Explorer.

Fully integrated set of different governance artifacts

As mentioned earlier, we provide business content across a wide range of different governance artifact types. Relationship Explorer fully supports all of these governance artifact types so it is possible to see the full array of integrations possible across different aspects of data governance.

In the example below, we can see how it is possible to view a comprehensive array of different governance artifact types all spanning out from one of our Business Terms (in this case : Social Security Number) . With this we can see the interconnections across Rules, Policies, Data Classes, Classifications as well as any relationships to relevant Tables and Columns.

Linking to Logical Data Models

Finally , it is also possible to import Data Models into IBM Knowledge Catalog and view them with Relationship Explorer. We provide pre-defined integration between the Knowledge Accelerators and the equivalent legacy Industry Models , for example the mappings between the current Knowledge Accelerator for Healthcare and the previous Unified Data Model for Healthcare( UDMH) as shown in the example below.

In this case, we can see how the different data model artifacts such as those from the Atomic Warehouse Model and the Dimensional Warehouse Model are connected to the healthcare term Patient. Of course, it is also possible to use Relationship Explorer as an effective way to visualize the relationships within a particular data model.

In summary, the availability of Relationship Explorer has graphically brought to life the rich network of business content in the various Knowledge Accelerators. It provides the user with the ability to build their own ad-hoc picture of artifacts relevant to their current governance scope. Indeed, any organization who is building out a comprehensive layer of business and technical metadata in IBM Knowledge Catalog will soon come to rely heavily on this critical new feature.

For more information about the Relationship Explorer feature in general, check out this blog https://medium.com/@michaszylar/simplifying-data-governance-and-boosting-data-literacy-with-ibms-relationship-explorer-050634ccfb8b

--

--

Pat O'Sullivan

Senior Technical Staff Member with IBM. A Data Architect with a background in Data Models, Business Glossaries, Data Governance and Data Management.