Production Rules on Databases
summary
A production rules engine called Naga rule engine implements a Rete-style variant on graph databases and runs over RDF and Datomic data, importing/exporting JSON and exposing an API plus a CLI. It purposely remains monotonic-only (no deletions) for simpler, auditable semantics and uses semi‑naive counting to avoid full re-evaluation. Ontology shortcuts reduce dozens of rules to one, and a working Datomic adapter exists; the system is open‑source and is already used at Cisco Threat Grid.
AI-written summary · about
chapters (17)
- 0:00Introduction to logic
- 0:35Understanding predicate logic
- 1:45Representing data relationships
- 2:42Rules and horn clauses
- 5:01Datalog and its restrictions
- 7:28Database reasoning techniques
- 9:18Rule execution example
- 11:28The Rete algorithm
- 14:10The Naga rule system
- 16:22Graph database principles
- 19:41Implementing rules with Naga
- 22:12Rule dependency management
- 26:52Handling negations
- 28:31Ontological modeling
- 31:27Advanced Naga features
- 34:21Naga demonstration
- 39:43Future plans and conclusion
related talks
41mDatomic Auto Schemas
23mNot Your Mothers Datalog
01h 00mFrom Planning To Production-Ready RAG
52mData Testing with HEDDA.IO
33mLogic at the Core: Why Business Rules Belong in the MarkLogic Platform
35mIt's Just Data
41mMerging the relational, the graph and the semantic
01h 06mDesign and Implementation of the RelationalAI Knowledge Graph Management System