AI systems are information systems.

Why this site exists

Most writing about AI is either a research paper or a product announcement. Neither helps the engineer who has to decide, this quarter, how their company’s knowledge should be structured so that a model can use it without inventing things.

Information science has answered versions of that question for decades. Ontologies, controlled vocabularies, provenance, and data modeling are mature disciplines. The guides here bring them to the current stack of models, retrieval systems, and agents, and keep the vocabulary consistent from one guide to the next.

How the guides are written

• Every guide places its subject on the same layered reference architecture.
• Concepts are defined before tools are named, so the material outlives any vendor.
• Diagrams explain a mechanism. They are never decoration.
• Each guide states the questions it answers and the ones it leaves to another guide.
• Guides are updated when the field moves. Updates are included in the original purchase.

Who writes them

The guides are written and edited by practitioners who have built knowledge systems, data platforms, and AI products in production, and reviewed by people who have to use the result.

Contact

Questions, corrections, and requests for new guides go to hello@informationscience.ai.