Causal Semantic Diagram Generator

Overview

The Causal Semantic Diagram Generator is a Hellixia feature that produces a Causal Semantic Diagram, i.e., a graph where nodes represent entities and arcs specifically represent causal relationships identified by Hellixia.

The generator queries a Large Language Model (LLM) to extract entities as nodes and the causal relationships between them as directed arcs. The result is a semantic graph that lays out the qualitative causal narrative of a domain.

Think of it as a tool that maps out “what causes what” in a domain, but without quantifying the strength of those effects. This is an unquantified diagram: it captures the existence and direction of cause-effect links, but not their magnitude. This makes it useful for rapidly mapping the causal structure of a topic before any data or parameters are introduced.

Usage

This is a node-centric workflow that lets you “grow” a diagram outward from a chosen seed concept.

You can optionally provide a Knowledge File for additional context.

Key Distinction

A Causal Semantic Diagram is not a Causal Bayesian Network. The latter is quantified with Conditional Probability Tables and supports causal inference (estimating effect sizes). The Causal Semantic Diagram is the unquantified, structural precursor, especially useful for mapping the causal landscape before moving to probabilistic modeling.

Workflow

Related Functions

• Causal Semantic Diagram Generator

  →
• Causal Relationship Finder

  →