Knowledge Graph and Master Data Management

My friend Henrik Gabs Liliendahl wrote a blog on MDM and Knowledge Graph that piqued my interest in the blogosphere. This blog is my take on the current state of knowledge graph used in master data management.

The knowledge graph is a digital structure representing knowledge as concepts and their relationships as facts. Knowledge graphs often store interlinked descriptions of entities – objects, events, situations, or abstract ideas. A knowledge graph can include an ontology that allows both humans and machines to understand and reason about its contents.

Knowledge graphs are often associated with linked open data projects, focusing on the connections between concepts and entities. An excellent example of this is how the search engines such as Google, Bing, and Yahoo work. Google uses a knowledge graph to enhance its search engine’s results with information gathered from various sources. Google users see the information in an infobox next to the search results. You can also see the applications of knowledge graphs in question-answering services such as WolframAlpha, Apple’s Siri, and Amazon Alexa, and social networks such as LinkedIn and Facebook.

Knowledge Graph and MDM

As Henrik points out, most attempts to combine MDM and knowledge graphs have been to visualize MDM relationships using a graph presentation. These relationships vary in nature. Some common examples are

Person-to-person relationships. These relationships are common in a household type of relationship spanning spouses, children, brothers, sisters, parents, etc.
Person-to-Business relationships. These include employee, contractor, agent, owner, CxO, etc., with a business entity.
Business-to-Business relationships. These types of relationships represent parent-subsidiary relationships, OEM relationships, company-supplier relationships, corporate partnerships, financial and legal relationships, competitors, etc.

While organizations can manage some of these relationships in a relational database structure, the limitations creep out as the volume and complexity increase. For instance, a relational database can efficiently capture my relationship with my wife and daughter. It will start to break when you need to grab all my friends (I have 700+ friends on Facebook, 13000+ followers on Twitter, and 5400+ connections on LinkedIn) and understand how they are related to me, as explained in this blog on social media and MDM. Similarly, business-to-business relationships can get complex if you want to manage all the business’s customers, competitors, and subsidiaries.

Knowledge graphs help make information retrieval more precise and help uncover previously unknown relationships in large data sets. While they have clear advantages to managing MDM relationships, I have not seen many organizations explore this area. While many vendors have added graph capabilities to their MDM offering, I do not believe the usage has lived up to the hype.

I have seen specific industries, such as pharmaceuticals, taking advantage of this technology. In pharma, doctors significantly influence the payer and provider network. A knowledge graph can help sales teams understand real-world affiliations better to tap into the most influential doctors. I also believe there is a strong case for a knowledge graph for product data. Suppose we can manage product relationships such as related products, competitive products, etc., in a knowledge graph, we can feed marketing, advertisement, search ranking, and recommendation to a great degree to fuel e-commerce. I discussed product mastering in this blog.

I am eager to hear what you think. Please share your feedback via comments or reach out to me on Twitter at @MDMGeek.

COMMENTS

No Thoughts on Knowledge Graph and Master Data Management

MDM – A GEEK'S POINT OF VIEW