Glossary/Knowledge Representation

Entity

An Entity is a distinct object or concept that can be uniquely identified and described using properties and relationships, serving as a fundamental unit in knowledge representation and data modeling.

Entities are the nouns of information systems: customers, products, accounts, events. Each entity is uniquely identifiable (customer #1234) and has properties (name, email, address) that describe it. Entities can participate in relationships with other entities: a customer places orders, an order contains products, a product is manufactured by a supplier. This entity-relationship model is foundational to both relational databases and graph-based knowledge representation.

The distinction between entities and their properties matters for analytics and AI. A customer is an entity; their email address is a property. A location is an entity; its latitude/longitude are properties. However, this distinction can be flexible: depending on context, a "location" might be treated as a property of a customer or as an entity in its own right with relationships to customers, warehouses, and sales regions.

Entity identification is critical for data integration and AI. Systems must recognize that "John Smith" in one database and "J. Smith" in another refer to the same entity. This process, called entity resolution or matching, is complex and essential for analytics where duplicate entities distort metrics. Knowledge graphs and AI systems reason about entities and their properties, making entity definition foundational.

Key Characteristics

  • Uniquely identifiable within a scope or using a unique identifier
  • Possess properties that describe characteristics or attributes
  • Participate in relationships with other entities
  • Can be of different types (customer, product, event, location)
  • Maintain identity even as properties change
  • Can be organized into hierarchies (all customers are entities, but categories exist)

Why It Matters

  • Provides foundational structure for data models and knowledge representation
  • Enables entity resolution where systems recognize the same entity across sources
  • Facilitates relationship analysis by making entity connections explicit
  • Supports AI reasoning about objects and their properties
  • Provides a natural way to understand and query information
  • Enables dimension modeling in analytics where entities are dimensions

Example

In a customer analytics system, a Customer entity has properties (name, email, segment), is uniquely identified by a customer ID, and participates in relationships (places Order, has Account, belongs to Region). Recognizing customer entities across systems (CRM, accounting, support) and matching them accurately enables unified customer analytics.

Coginiti Perspective

SMDL entities map directly to tables or SQL queries, forming the foundation of Coginiti's semantic model, where each entity's dimensions and measures define the properties available for analysis. Entity relationships defined in SMDL enable Semantic SQL to perform implicit joins, ensuring consistent entity interpretation across all queries and tools connected through the ODBC driver.

See Semantic Intelligence in Action

Coginiti operationalizes business meaning across your entire data estate.