Propositional Causal Bayesian Network Generator

Overview

The Propositional Causal Bayesian Network Generator is Hellixia’s most powerful feature for automatically building a fully specified Causal Bayesian Network that explains a selected target node. It is accessed via the Hellixia menu after designating a target node in your BayesiaLab graph.

Functionality

Causal Discovery

Hellixia queries a Large Language Model (LLM) to identify the causal ancestors (and optionally descendants) of the target node. The LLM also assigns an estimated direct effect, a numerical value between ‑1 (negative influence) and +1 (positive influence).

Structure & Parameterization

The identified nodes and arcs are added to the graph. Conditional Probability Tables (CPTs) are automatically populated using the DualNoisyOr() and the SwitchNoisyOr() functions. The estimated direct effects serve as the per‑cause parameters of these functions.

Visual Encoding

Arcs are displayed according to the estimated causal strength: blue for positive influence, red for negative influence, with the line thickness reflecting the magnitude.

Comments

The generator also attaches comments to arcs explaining the underlying causal mechanism.

Usage

Ensure the target node already exists in your graph, with a meaningful Name, Long Name, and Comment, which helps the LLM understand what the node represents.

Set as Target Node: right‑click on the node or double‑click its status bar icon.

Then go to Hellixia > Propositional Causal Bayesian Network Generator.

Note that the generated network works only with Boolean nodes (True/False states) because the DualNoisyOr/SwitchNoisyOr functions are defined for binary variables.

Risk Causal Network Generator – A version specialized for risk modeling, using predefined role types (Trigger, Control, Event, Main Risk, etc.) and constraining the structure to a risk‑centric typology.

General Causal Network – A Constraint‑Based variant using the General typology (Event, Aggregate, Intervention, etc.), not risk‑focused.

Unconstrained Causal Network – Built by the Causal Relationships Finder or Multi‑Engine Causal Relationship Finder; does not force role constraints but still produces a quantified causal network.

Comparison with other Hellixia generators

Causal Semantic Diagram Generator – Produces a qualitative causal graph with arcs representing causal relationships, but no quantification (no CPTs, no estimated strengths). The Causal Network Generator goes further by fully specifying the probability distributions.

Knowledge Graph Generator – Creates a general semantic graph of entities and non‑causal relationships; the Causal Network Generator focuses exclusively on causal links and their quantification.

Example use case

The generator is often demonstrated on examples such as explaining Lung Cancer or Tuberculosis. For a detailed walkthrough, see the use case Generate a moderately complex Causal Bayesian Network to explain Lung Cancer