Data Modeling Glossary

data model

abstract model of data entities and relationships

A data model connects various domain vocabularies to “connect the dots” with a specific purpose in mind. It is intended to capture how the data are connected in ways that are meaningful to subject matter experts. Data models are specific to selected use cases or objectives. A good data model leverages standard vocabularies often in the form of ontologies and defines and extends concepts and relationships only when necessary. One can view a data model as a reusable design for one or more schemas.²

knowledge graph

graph-based model containing data

A knowledge graph is a data model graph populated with data.

ontology

vocabulary

An ontology is a formally specified vocabulary which defines the concepts and relationships that characterize a domain. Further, we assume that an ontology represents all key high-level concepts in a domain and minimally provides parent-child relationships among the concepts.

schema

format for data

A schema defines the structure generally used to store or transmit data.

Examples of schemas include HCA's JSON schemas, a Postgres or MySQL schema, or a schema for Google's BigQuery.

A schema is often an implementation of a data model either implicitly or explicitly but usually includes optimizations for storage or performance that are not relevant to the data model.