BOOMER-PY

BOOMER-PY (Bayesian OWL Ontology MErgER in Python) is a probabilistic reasoning system for knowledge representation and ontological reasoning with uncertainty.

Overview

BOOMER-PY enables reasoning over probabilistic facts and taxonomic relationships, finding the most likely consistent interpretation of potentially conflicting assertions. It uses a combination of graph-based reasoning and Bayesian probabilistic inference.

Key features:

• Represent probabilistic ontological statements
• Reason over class subsumption hierarchies
• Evaluate class equivalence relationships
• Detect and resolve logical inconsistencies
• Calculate posterior probabilities for each assertion

Core Concepts

• Knowledge Base (KB): Collection of facts and probabilistic facts (PFacts)
• Facts: Logical assertions about entity relationships
  • SubClassOf: A is a subclass of B
  • ProperSubClassOf: A is a proper subclass of B (A � B)
  • EquivalentTo: A is equivalent to B
  • NotInSubsumptionWith: A is not in a subsumption relationship with B
  • MemberOfDisjointGroup: A belongs to disjoint group G
• Probabilistic Facts: Facts with assigned probabilities
• Reasoning: Logical deduction over facts to find satisfiable solutions
• Search: Exploration of possible combinations of assertions

Use Cases

BOOMER-PY is designed for:

• Merging ontologies with uncertain mapping relationships
• Reasoning with probabilistic taxonomies
• Resolving conflicts in knowledge bases
• Scientific knowledge representation with uncertainty

Example

from boomer.model import KB, PFact, EquivalentTo
from boomer.search import solve
from boomer.renderers.markdown_renderer import MarkdownRenderer

# Create a knowledge base with probabilistic facts
kb = KB(
    pfacts=[
        PFact(EquivalentTo("cat", "Felix"), 0.9),
        PFact(EquivalentTo("dog", "Canus"), 0.9),
        PFact(EquivalentTo("cat", "Canus"), 0.1),
    ]
)

# Solve to find most probable consistent solution
solution = solve(kb)

# Display results
renderer = MarkdownRenderer()
print(renderer.render(solution))

Installation

# Clone the repository
git clone https://github.com/your-username/boomer-py.git
cd boomer-py

# Install dependencies
pip install .

Development

BOOMER-PY uses:

• NetworkX for graph-based reasoning
• Pydantic for data modeling
• Pytest for testing

To run tests:

make test