The Fall and Rise of Organizational Ontologies

The Fall and Rise of

Organizational Ontologies

By Ross A. McIntyre

Originally posted on LinkedIn, 2023

The rise of applied artificial intelligence has prompted a resurgence in the consideration of ontologies as they relate to business organizations. Organizational ontologies refer to structured frameworks or sets of concepts that represent knowledge about an organization. Ontologies are back – even if you never noticed that they left.

The concept of an ontology has deep roots in philosophy and its origin can be traced back to the teachings of Plato and Aristotle. Ontologies concern fundamental questions such as "What is reality?", "What is the relationship between mind and body?", and "What is the nature of time and space?" Aristotle argued for multiple categories of being: substance, quality, quantity, and relation. David Hume offered a perspective that knowledge is based on experience and there is no such thing as a necessary truth. Immanuel Kant posited that the mind has a priori categories that structure our experience of the world and those are divorced from direct experience. Philosophical and organizational ontologies are related but complementary fields of study that can be used to improve the way that organizations understand and manage their information.

Ontologies are back – even if you never noticed that they left.

Tim Berners-Lee, the creator of the World Wide Web, has been a strong proponent for ontologies in the Semantic Web. He holds the position that ontologies are essential for creating a web of machine-readable data and, in fact, are the key to unlocking the full potential of the web. In a 2001 article in Scientific American, he wrote:

“The Semantic Web (aka Web 3.0 – ed) is not a separate web but an extension of the current web that is based on machine-understandable information. This means that not only can computers index and retrieve information from the web, they can also reason about it and use it to solve problems.”

So, what does that have to do with me and how did you get in here, you ask? Applying philosophical ontologies to businesses can aid in the creation of organizational ontologies in a variety of ways: clarifying the meaning of concepts, identifying different organizational entities, developing taxonomies, and designing information systems. In turn, companies may benefit through improved communication, better decision-making, and enhanced customer satisfaction. Organizational ontologies, for instance, concern the hierarchy of roles, the relationship between departments, understanding of workflows, and knowledge of services, products, and offerings. With the rise of A.I. and machine learning technologies, ontologies have become critical to navigating mass data volume, automating processes, and aiding in decision-making. Here are a few concepts that benefit from such ontologies:

However, truly leveraging the potential of data requires humans to define rules and relationships that govern an organization’s operation, i.e., an ontology. Nick Vrana , Stellar Elements Global Director of Telco and AI Platforms put it thusly:

“Interestingly, we’ve tried all of these things over and over again, not because they are (self-aggrandizing), but…because they would be useful if we could get them to work for everybody. Organizational ontologies have always been a problem, for example, because they are, by nature, inflexible… the world is not perfectly described by language. But AI…is good at inferring or dealing with the grey and helping to connect it to our black-and-white.”

Numerous businesses are already utilizing ontologies to manage organizational information. For instance:

When infinite complexity smashes into our limited descriptive ability, the only way that can be handled, according to Vrana, is “with errors and humans.”

All our previous attempts to put structure to chaos have failed because the structure has not been flexible enough to deal with the real world. Now, we might have tools that can provide the “boundary layer” between infinite complexity and the indescribable nature of the real world and our digital systems which, by nature, need things to be more strictly defined. The existence of heretofore indefinable things is not a limitation of humanity, nor a limitation of technology, it is a limitation of physics – we can add infinite terms to the dictionary to approximately describe the infinitely complex. Currently, when infinite complexity smashes into our limited descriptive ability, the only way that can be handled, according to Vrana, is “with errors and humans.” An AI can be an intermediary to those two things and infer or “connect the dots” between a computer system’s model of the world to figure out what is trying to be accomplished.

Improved Data Management: Advanced AI and machine learning algorithms need structured data to function properly. Ontologies offer a structured way to tag, organize, and interpret data, especially in large organizations with multiple functions and complex operational structures.
AI and Automation: Ontologies help guide AI algorithms to better understand and interpret organizational operations, which is essential in automating tasks and allowing AI to work more effectively alongside human employees.
Interoperability: As organizations use more and more disparate systems and tools, ontologies provide a shared vocabulary, ensuring all systems can speak to each other and work together seamlessly.
Decision-making: With a clear representation of the organization's structure and operations, decision-makers can make more informed decisions based on the relationships and processes outlined in the ontology.
Enhanced Search and Discovery: In an era where organizations manage vast amounts of data, ontologies can facilitate more effective information search and discovery by providing structure and semantic understanding.
Improved Regulatory Compliance: Organizations are increasingly subject to regulations that require them to manage and share data in a certain way. Ontologies can provide a way to structure and manage organizational data in a way that is compliant with regulations.

IBM: IBM uses ontologies to support a wide range of applications, including product development, supply chain management, and customer relationship management. IBM's Watson system, which is an AI platform, uses ontologies to understand the meaning of natural language and to answer questions in a comprehensive and informative way.
Siemens: Siemens manages its knowledge base, which includes over 100 million documents, with ontologies. Siemens' knowledge base is used to support a wide range of activities, including product development, maintenance, and customer support.
Boeing: Boeing uses ontologies to manage the design and manufacturing of its airplanes. Boeing's ontologies help to ensure that different parts of the airplane are designed and manufactured in a way that is compatible and safe.
Johnson & Johnson: Managing clinical trials data is facilitated via ontologies at Johnson & Johnson. Their ontologies help to ensure that the data is accurate, complete, and consistent.
BP: BP uses ontologies to manage its oil and gas exploration and production data. BP's ontologies help to ensure that the data is used effectively to make decisions about where to explore and produce oil and gas.