The Role of the Ontology

What is an Ontology?

Ontologies are frequently mentioned in the context of knowledge graphs. Ultimately, an ontology serves to create a formal representation of the entities in a graph. They are usually based on a taxonomy, but since they can contain multiple taxonomies, it maintains its own separate definition.

An ontology distinguishes between the different concepts in a domain and defines how they can be related. For example, if we examine a particular venue, like Madison Square Garden, an ontology distinguishes between the events at that location using a variable such as time. A sports team, like the New York Rangers, has a series of games within a season that will be hosted in that arena. They are all hockey games, and they are all located in the same venue. However, each event is distinguished by its date and time.

Your Schema for Extraction

In the Perseus Client, the Ontology is the most important control mechanism in the Text-to-Graph pipeline. It is the schema that defines what to extract from your text and how to structure the resulting Knowledge Graph.

Think of it like a blueprint for a building or a schema for a database. It provides the rules and vocabulary for your domain, ensuring that the output is predictable, clean, and tailored to your specific needs.

What an Ontology Defines

At its core, an ontology defines the classes of "things" that can exist in your graph, how they can relate to one another, and their associated characteristics. Specifically, it defines:

• Entity Types: The allowed categories for your entities. For example, you might define Company, Person, and Investment as valid entity types.
• Relation Types: The allowed connections between those entity types. For example, you could define that a Company can have an INVESTED_IN relation with another Company.
• Properties: The attributes or characteristics that entities and relations can possess. For example, a Person entity might have properties like name, age, or email.

Why It Matters

Using an ontology-driven approach provides several key advantages:

• Consistency: It ensures that all extracted information conforms to a single, unified model.
• Relevance: It focuses the extraction process only on the concepts that matter to you, reducing noise from irrelevant information.
• Clarity: It makes the resulting Knowledge Graph easy to understand and query because the structure is explicit and well-defined.
• Power: It allows you to build highly domain-specific AI systems that understand the nuances of your industry or organization.

You can either upload a pre-existing ontology or use the Console to build a new one from a source document, providing a use case and language to guide the generation process.