日本語 
Big computer programs that understand a lot of language have changed how companies think about intelligence. These programs can do things like summarize papers, write code answer questions and work with an amount of information that is not organized. But now that companies are using intelligence to make decisions and run their businesses they are also looking at something else: ontology.
There was a talk about this on AI Exploration Journey. They said ontology is becoming more important because big language models need to understand how things are related to each other. This means understanding how information, people and business ideas fit together. For a country like Japan, where companies often have systems and do things in special ways this is a big deal.
Ontology is like a map that shows how concepts are connected. It is not a list of things but a way to understand what they mean and how they relate to each other. This makes it useful, for helping artificial intelligence programs understand the information that companies have and for making sure that the programs are using the information to make decisions. Artificial intelligence and ontology are related because ontology helps artificial intelligence understand the meaning of things and artificial intelligence can use ontology to make decisions.
Why Large Language Models Are Bringing Back the Concept of Ontology
The fast growth of intelligence tools has shown a problem that was sometimes not obvious because of exciting examples. A large language model can understand words and sentences well but it does not automatically know the specific rules, the special words and the connections inside a companys work.
Think about a company that has different systems for buying things making products keeping track of stock and checking quality. Each system might use words and numbers. An artificial intelligence helper can get information, from those systems. Without a shared way of understanding meaning putting the information together correctly can be hard.
Ontology can provide that missing structure
By defining concepts such as products, suppliers, machines, orders and production processes—and specifying how they relate—businesses can give AI systems a more consistent representation of their operational environment. Recent research is also exploring how LLMs can assist with creating and maintaining enterprise ontologies, potentially reducing some of the traditionally manual work involved.
Japan Could Benefit From the Shift
Japan’s technology industry has a particularly strong reason to pay attention to this development. Many Japanese companies have accumulated decades of business data across legacy applications, spreadsheets and departmental databases.
This is where ontology-based artificial intelligence could really be helpful. Businesses do not just need to link a language model to a bunch of documents. They can create a system that explains what their data is about and how all the different parts of information are related to each other. This system is, like a foundation that helps businesses understand their ontology-based intelligence and data.
For Japanese manufacturers, this could support applications involving production planning, equipment maintenance and supply-chain management. Financial institutions could use semantic models to connect customer, transaction and regulatory information. Healthcare organizations could benefit from clearer relationships between patients, treatments, medicines and clinical records.
The potential value comes from making AI more context-aware rather than simply making the underlying language model larger.
Implications for Japan’s AI and IT Services Market
The return of ontology could create a new market opportunity for Japanese technology companies.
Japan’s IT services industry has traditionally played a major role in integrating complex enterprise systems. As businesses begin adding AI agents to those environments, demand could shift toward companies capable of combining data engineering, knowledge graphs, ontology development and AI implementation.
This is where ontology-based artificial intelligence could become really valuable. Businesses do not just need to link a language model to a bunch of documents. They can create a foundation that explains what their data means and how the different pieces of information are connected to each other. This way ontology-based artificial intelligence helps businesses understand their data better.
Businesses Could Gain More Reliable AI
For businesses, the biggest attraction may be reliability.
An AI system that has access to company documents can retrieve relevant information, but retrieval alone does not guarantee that the system understands the relationships between pieces of information. Ontology-based approaches attempt to add that missing layer of structure.
This could be especially useful for AI agents. As businesses move toward agents that can execute tasks rather than simply answer questions, errors caused by misunderstanding business rules become more costly.
An agent managing procurement, for example, needs to understand not only what a supplier is but also approval limits, product relationships, contractual conditions and organizational responsibilities.
That is a very different requirement from generating a well-written paragraph.
A New Role for AI Engineers and Data Specialists
The growing interest in ontology could also affect technology employment in Japan.
Companies might need people who know both ways of building knowledge systems and new artificial intelligence. Data designers, software programmers and AI creators could work with people who know areas well to set up the ideas and connections that are important for company AI systems.
It’s interesting that large language models could help lower the cost of this work. Scientists are looking into using AI to help build ontologies, where language models find ideas and connections while people with experience still check and manage everything.
This way of working could make it easier for smaller Japanese companies to create knowledge systems. They didn’t have the money or tools before to build these kinds of systems.
The Road Ahead for Enterprise AI in Japan
The new interest in ontologies doesn’t mean that large language models are less important. The opposite is true. It means that the next group of AI used by companies will rely on mixing language models with data systems.
Also Read: Oracle and AWS Drive Growth in Enterprise AI Databases
Japanese businesses have a chance to use this method to link their knowledge, their programs and their daily data, with new AI helpers. The outcome could be systems that do more than write words—they could understand business situations follow company rules and help make decisions using organized company knowledge.
For Japan’s technology industry, that represents an important change in priorities. The next AI advantage may not come solely from having access to the newest model. It may come from having the best understanding of the business data surrounding that model.
As generative AI moves deeper into Japanese enterprises, ontology could therefore evolve from an overlooked information-science concept into an important component of the country’s enterprise AI architecture.
English