Ontologies
Formal representations of knowledge defining concepts, properties, and relationships within a domain, used to structure information for AI.
Ontologies are like detailed maps of knowledge. They define specific terms and explain how those terms relate to each other within a particular subject. Think of it like a dictionary and an encyclopedia combined, but structured so a computer can understand it.
For example, an ontology for a library might define terms like 'book,' 'author,' and 'genre.' It would also specify relationships, such as 'a book has an author' and 'a book belongs to a genre.' This structured knowledge helps AI systems understand and reason about the information they process.
Ontologies provide a formal, explicit specification of a conceptualization. They consist of classes (concepts), properties (attributes and relationships), and axioms (rules and constraints) that define a domain of interest. Ontologies enable knowledge sharing and reuse by providing a common vocabulary and semantic framework.
In AI, ontologies are used to ground language models in factual knowledge, improve reasoning capabilities, and facilitate knowledge graph construction. They allow for more precise querying and inference over data, moving beyond simple keyword matching to semantic understanding. Tradeoffs include the significant effort required for ontology development and maintenance, and the potential for rigidity if not designed flexibly.
As AI agents become more integrated into business workflows, the need for structured knowledge representation is growing. Ontologies help ensure AI agents can accurately understand and utilize complex business data, leading to more reliable and grounded responses. This is crucial for applications where precision and domain-specific understanding are paramount.
- Your Agent Is Only as Smart as Its ContextMicrosoft · Microsoft Developer on YouTube ·
- Microsoft Copilot
- Microsoft Copilot leverages ontologies to connect AI agents with business data, enabling them to provide accurate and contextually relevant information.
The term 'ontology' in the context of information science and AI originates from philosophy, where it studies the nature of being. In computer science, it was popularized in the AI community in the 1990s for building knowledge-based systems. Source ↗
You can explore ontologies by looking at public knowledge graphs like Wikidata or DBpedia. Try searching for a concept, like 'Paris,' and observe how its properties (e.g., capital of, population) and relationships to other entities are presented. This demonstrates how structured knowledge is organized.
The Playbook: reusable how-tos, prompts, configs and tips pulled out of the news.
- PromptRole-play a student struggling with math for AI tutor simulationSimulate a student's math struggles and uncertainty for AI tutor evaluation. · 5 min
- TemplateReusable workflow templates for healthcare teamsTeams can create and share templates for common workflows like documentation or prior authorizations. · 15 min
- TemplateTemplate for building an AI tool tailored to your workA structured approach to defining and building a custom AI tool for specific business needs. · 15 min
- TemplateEvals CSV prompt list for agent skillsA CSV file template for listing prompts used in evaluating agent skills. · 20 min
- HeuristicDouble coding agent task completion with docs linksProviding coding agents with a documentation link significantly increases their task completion rate. · 2 min
Questions people ask
- What is the main goal of an ontology?
- The main goal is to create a shared understanding of a domain that can be used by both humans and machines, enabling better data integration and reasoning.
- How are ontologies different from simple databases?
- Ontologies define the meaning and relationships between data elements, allowing for more complex queries and inferences than a typical database which primarily stores structured data.
First seen on Subvolts Oct 6, 2026. Updated Oct 13, 2026. Written by an AI model from the items above and reviewed against them weekly; report an error.