GraphNews

Knowledge Graph schema — when you're building a system to track contributions at scale. Claude generates a complete graph schema for Neo4j, Postgres, or JSON-LD. Nodes for code artifacts, human sources, individual experts, AI sessions, and knowledge claims. Edges that let you ask: "who are the humans behind this module?" or "what did @username contribute to this codebase?"

Holons, Encapsulation, Reification, and the REST Principle in Ontology Design

(And What To Do About It)

The Cagle Report | Kurt Cagle We have spent thirty years building better and better databases. Relational tables, document stores, graph databases, data lakes, knowledge graphs — each generation has improved our ability to capture what things are.

Bridging the gap between organisational data, implicit organisational reasoning, and AI reasoning with Knowledge Graphs purpose-built for AI.

From Agile to Design Thinking. Scrum assumes tame problems that can be time-boxed, estimated with planning poker, decomposed into user stories, tracked on a velocity chart and declared done against a checklist. If only knowledge were this tame. Knowledge organization is anything but tame.

Retrieval-Augmented Generation (RAG) systems typically treat documents as flat text, ignoring the structured metadata and linked relationships that knowledge graphs provide. In this paper, we...

Web de données | Architecture de l'informations | Accès aux connaissances

Web de données | Architecture de l'informations | Accès aux connaissances

The Role of Competency Questions in Keeping Ontology Modeling on Track Competency questions (CQs) are the compass and the yardstick of ontology engineering. They are simple natural‑language questions an ontology must answer, helping teams stay focused on real user needs. Without them, projects drift into abstraction. With them, you keep a constant check — does the ontology answer meaningful questions? What Competency Questions Do Capture requirements: Turn vague goals into clear queries — e.g., “Which patients had a fever within 24 hours of receiving Drug X?” Define scope: Clarify what the ontology must and must not cover. Guide design: Each question implies which classes, properties, and relationships you need. Enable validation: Translate CQs into SPARQL; if they run successfully, the ontology works. Improve communication: Domain experts, managers, and engineers all understand CQs. How to Use Them Elicit from stakeholders: Collect 10–20 essential questions; keep them specific and relevant. Formalize: Write each as a SPARQL template — missing elements reveal modeling gaps. Prioritize: Start with top 5–10, then expand iteratively. Test continuously: Re‑run CQs after every ontology change to catch regressions early. Example: Drug Safety Ontology CQ: “List all patients with a serious adverse event within 30 days of starting Drug X.” → Forces modeling of Patient, Drug, AdverseEvent, and the relation occursWithin30DaysOf. As new CQs emerge, new elements (like ConcomitantMedication or overlapsWith) are added only when needed, avoiding under‑ or over‑engineering. Why It Matters CQs prevent scope creep, give measurable validation, and keep ontologies grounded in practical outcomes. They serve as both requirements and quality tests — ensuring every concept exists for a reason and every query finds its answer. The Bottom Line Competency questions are the backbone of purposeful ontology design — guiding requirements, shaping design, and validating success. In complex domains like life sciences, they are the surest way to stay aligned with real‑world needs. #KnowledgeGraphs #OntologyEngineering #KnowledgeManagement #SemanticAI #NeuroSymbolicAI #CausalAI #KnowledgeRepresentation #ExplainableAI #OntologyDesign #DigitalTransformation 👉 Follow me for Knowledge Management and Neuro Symbolic AI daily nuggets. 👉 Join my group for more insights and community discussions Join the Group https://lnkd.in/d9Z8-RQd

Measuring retrieval coverage, path faithfulness, provenance completeness, cost, latency, and decision usefulness instead of polished answers

Large Language Models (LLMs) exhibit strong abilities in natural language understanding and generation, yet they struggle with knowledge-intensive reasoning. Structured Knowledge Graphs (KGs) provide an effective form of external knowledge representation and have been widely used to enhance performance in classical Knowledge Base Question Answering (KBQA) tasks. However, performing precise multi-hop reasoning over KGs for complex queries remains highly challenging. Most existing approaches decompose the reasoning process into a sequence of isolated steps executed through a fixed pipeline. While effective to some extent, such designs constrain reasoning flexibility and fragment the overall decision process, often leading to incoherence and the loss of critical intermediate information from earlier steps. In this paper, we introduce KG-Reasoner, an end-to-end framework that integrates multi-step reasoning into a unified "thinking" phase of a Reasoning LLM. Through Reinforcement Learning (RL), the LLM is trained to internalize the KG traversal process, enabling it to dynamically explore reasoning paths, and perform backtracking when necessary. Experiments on eight multi-hop and knowledge-intensive reasoning benchmarks demonstrate that KG-Reasoner achieves competitive or superior performance compared to the state-of-the-art methods. Codes are available at the repository: https://github.com/Wangshuaiia/KG-Reasoner.

Why enterprise knowledge graph projects keep failing — and what actually fixes them By Kurt Cagle and Chloe Shannon | The Cagle Report Every few years, a large organisation decides it needs an enterprise ontology. A project is stood up.

I’ve spent the past year building GraphRAG systems from scratch. Here’s the architecture I keep coming back to (steal it)… MongoDB is the unified memory. Voyage AI by MongoDB for embeddings. Agent on top. Here's how the end-to-end flow looks: 𝟭–𝟮 / 𝗜𝗻𝗴𝗲𝘀𝘁𝗶𝗼𝗻 → 𝗗𝗮𝘁𝗮 𝘄𝗮𝗿𝗲𝗵𝗼𝘂𝘀𝗲 Pull data from URIs, notes, emails, docs. Normalize into a single document schema and store in MongoDB. Example: emails, research notes, and meeting transcripts become raw_documents. This is the durable ingestion layer. 𝟯–𝟳 / 𝗠𝗲𝗺𝗼𝗿𝘆 𝗽𝗶𝗽𝗲𝗹𝗶𝗻𝗲 Each document flows through: • Clean text + metadata • Chunk (optional) • Graph extraction (entities + relationships) • Normalization (merge duplicates like “Abi” vs “Abi Aryan”) • Embeddings with Voyage AI Output: knowledge graph objects with triplets, vectors, and metadata. 𝟴 / 𝗨𝗻𝗶𝗳𝗶𝗲𝗱 𝗺𝗲𝗺𝗼𝗿𝘆 𝗶𝗻 MongoDB Materialize into a knowledge_graph collection with entities and relationships as documents. Indexes: • Text index → keyword recall • Vector index → semantic recall • Graph links → multi-hop traversal Documents, vectors, and graph memory in one place. 𝟵 / 𝗠𝗖𝗣 𝘀𝗲𝗿𝘃𝗲𝗿 Expose GraphRAG through tools: • NL query → compact high-level retrieval • Deep search → progressive graph expansion • Ingest → URLs, files, conversations 𝟭𝟬 / 𝗛𝗮𝗿𝗻𝗲𝘀𝘀 Claude Code orchestrates reasoning and tool usage. It decides when to retrieve vs write memory. 𝟭𝟭 / 𝗦𝗸𝗶𝗹𝗹𝘀 Skills define interaction with GraphRAG: • Assistant-memory → harness to MCP bridge • Assistant-learn → push insights back to memory 𝟭𝟮–𝟭𝟯 / 𝗔𝗴𝗲𝗻𝘁𝗶𝗰 𝗚𝗿𝗮𝗽𝗵𝗥𝗔𝗚 The agent selects tools dynamically. Semantic + text search retrieves entry nodes. Graph traversal (2–3 hops) expands context. Example: "Create GraphRAG talk for O’Reilly" → Retrieve GraphRAG + O’Reilly → Expand to past talks and preferences → Return structured context MongoDB stores evolving graph memory. Voyage AI by MongoDB powers semantic recall. Agents turn retrieval into reasoning. One unified memory layer. One ingestion and retrieval pipeline. One agent loop. P.S. Are you building GraphRAG with a unified memory layer, or still splitting vectors, graphs, and documents across multiple databases? | 17 comments on LinkedIn

PDF | Graph-based Retrieval-Augmented Generation (GraphRAG) advances flat document retrieval by structuring knowledge as relational graphs, enabling... | Find, read and cite all the research you need on ResearchGate

The Governance Layer Enterprise AI Has Been Missing

Vector search fails predictably on multi-hop reasoning queries. GraphRAG addresses that gap — but introduces a different cost structure, failure modes, and maintenance burden that most teams underestimate.

BEIJING, April 10, 2026 /PRNewswire/ -- As the way of managing enterprise data assets evolves from simple accumulation to value extraction, the role of AI has shifted accordingly: it is no longer limi

Building a Fully Agentic Employee Knowledge Graph: OrgGraph AI — using LangGraph & Neo4j Today’s global enterprise faces a massive talent visibility problem. We operate on a massive paradox: we …

When working with large interconnected code repositories, especially once with microservices that aren't obvious how they're related, AI coding...

How moving from Word docs to network graphs transformed my research workflow and how LLMs made it possible.