https://www.heuristicontology.com/
The Heuristic Decision Ontology provides a structured, semantically rigorous framework for modeling cognitive heuristics as directive information content entities, grounded in the principles of Basic Formal Ontology (BFO 2020) and Common Core Ontologies (CCO 2.0). It systematically organizes heuristics - simplified cognitive strategies used to reduce mental effort and enable fast, satisficing decisions under uncertainty or constraint - into clearly defined classes and hierarchical structures.
Each heuristic is modeled not merely as a psychological tendency but as a prescriptive specification that can guide agent behavior. The ontology supports the explicit representation of heuristic-guided processes and is designed to allow users to define object properties that link these processes to the heuristics that inform them. While such relationships are not currently implemented in the ontology, they are anticipated in future versions to support computational reasoning, behavioral simulation, and integration with intelligent systems and cognitive architectures.
By treating heuristics as reusable, interoperable components within broader decision-support ontologies, the framework enables modeling of real-world behaviors in domains such as cybersecurity, medical triage, mission planning, and user-interface design. It supports traceability of decision strategies, analysis of context-specific cognitive adaptation, and the evaluation of alternative heuristics under varying operational constraints.
The Heuristic Decision Ontology provides a formal semantic framework for representing heuristics as directive information content entities - structured, named, and computationally actionable constructs. Unlike informal conceptions of heuristics as vague or unconscious tendencies, this ontology treats them as clearly defined, prescriptive specifications that can be operationalized by intelligent agents to guide decision-making. Heuristics are understood not as psychological traits but as modular information structures that direct behavior in the face of uncertainty, limited information, or time constraints. Grounded in Basic Formal Ontology (BFO 2020) and aligned with the Common Core Ontologies (CCO 2.0), this ontology ensures conceptual clarity, formal rigor, and interoperability across knowledge systems and application domains.
This ontology is designed to accompany the paper and presentation at the Proceedings of the Joint Ontology Workshops (JOWO) – Episode XI: The Sicilian Summer under the Etna, co-located with the 15th International Conference on Formal Ontology in Information Systems (FOIS 2025) on September 8-9, 2025.
This ontology is undergoing active development. An updated and expanded version should be available shortly. Please check the repo for the most updated version or contact the creator.
The Heuristic Ontology models cognitive heuristics as Directive Information Content Entities (DICE), enabling formal representation of decision-making under constraints such as time pressure, incomplete data, or limited computational capacity. It organizes heuristics into four cognitive families—Decision Simplification, Exploratory, Social, and Temporal—each with specific subclasses (e.g., Availability Heuristic, Social Proof Heuristic). This version enhances support for medical triage applications, adding new properties, classes, instances, and reasoning constraints to model scenarios like the Simple Triage and Rapid Treatment (START) protocol.
Developed by researchers at SUNY University at Buffalo and the National Center for Ontological Research, the ontology bridges cognitive science, behavioral economics, and applied AI. It supports use cases in medical triage, cognitive modeling, decision-support systems, and knowledge graph integration.
Classifying Heuristics into Cognitive Families: At the core of the ontology is the class Heuristic Instruction, a subclass of directive information content entity, which encompasses prescriptive guidance that simplifies cognitive effort by specifying what to attend to or how to act in a particular kind of situation. This class is subdivided into four top-level subclasses, each representing a distinct cognitive strategy family: Decision Simplification Heuristics, Exploratory Heuristics, Social Heuristics, and Temporal Heuristics.
Decision Simplification Heuristics prescribe strategies that reduce decision complexity by selectively focusing on salient, preselected, or cognitively accessible information. These heuristics enable rapid judgment by filtering or constraining input based on perceptual or contextual salience. Examples include the Default Effect Heuristic (favoring pre-selected options), the Availability Heuristic (basing judgments on the ease of memory retrieval), the Recognition Heuristic(favoring familiar options), the Representativeness Heuristic (relying on similarity to a prototype), the Anchoring Heuristic (relying on initial reference values), and the Effort Heuristic (inferring value based on perceived effort invested).
Exploratory Heuristics guide behavior toward novelty, uncertainty resolution, or information gain. These heuristics support exploration when outcomes are uncertain or when learning is prioritized over exploitation. Examples include the Curiosity Heuristic, which directs attention toward incomplete or surprising information, the Novelty-Seeking Heuristic, which favors less familiar alternatives, and the First-Mover Heuristic, which guides action based on temporal order and the presumed advantages of acting early.
Social Heuristics rely on social context, cues, or norms to inform decision-making. These strategies simplify choices by deferring to trusted others, observed behaviors, or group signals. Examples include the Authority Heuristic (deference to experts), the Social Proof Heuristic (copying others' actions), the Trust Predisposition Heuristic (assuming others are trustworthy by default), the Ingroup Heuristic (favoring in-group perspectives), and the Reciprocity Heuristic (responding to favors or concessions with reciprocal action).
Temporal Heuristics prioritize or evaluate information based on its timing or temporal relevance. These strategies are activated in time-sensitive contexts where recency, sequencing, or deadline pressure shapes cognitive priorities. Examples include the Recency Heuristic, which emphasizes the most recent events or inputs, and the First-Mover Heuristic, when activated as a timing advantage rather than novelty preference.
Each of these top-level families is formally defined using genus–differentia criteria and includes a range of specialized subclasses. This typology supports extensibility, enables consistent classification, and facilitates reasoning over functional distinctions in cognitive strategy.
Semantic Grounding and Interoperability: Every heuristic in the ontology is modeled as a directive information content entity that can be named, annotated, and linked to specific classes of heuristic-guided processes. The formal structure supports integration with intelligent systems, agent-based models, cognitive architectures, and decision-support environments. Heuristics may be operationalized as components of broader planning structures, instantiated in interface design patterns, or embedded in knowledge graphs representing socio-technical systems. The use of formal ontological commitments ensures that heuristic strategies can be traced, justified, and reasoned over using standard OWL reasoning tools.
By encoding heuristics as ontologically grounded entities, this model bridges theoretical insight from cognitive science and behavioral economics with the technical requirements of applied AI and knowledge engineering. The Heuristic Ontology enables the representation of bounded rationality, satisfying behavior, and domain-specific cognitive strategies in a way that is reusable, extensible, and machine-interpretable. It provides a foundation for understanding how human and artificial agents make decisions under constraint - and for building systems that model, augment, or ethically intervene in those behaviors.
- Version IRI: http://www.heuristicontology.com/2025-08-16/HeuristicDecisionOntology
- Release Date: August 18, 2025
- Changes in This Version:
- New Object Properties:
guided_by,guides_behavior,has_intended_outcome,is_applied_to,has_participant,has_input,has_outputto link heuristics to processes and outcomes. - Pending Data Properties:
confidenceLevel,cognitiveLoad,responseTime,expectedAccuracy,learningCurve,applicabilityScoreto quantify heuristic performance. - Pending Medical Triage Classes:
MedicalTriageHeuristic,STARTTriageHeuristic,MedicalTriageDecisionProcessfor triage-specific modeling. - Pending Instances: Added instances for triage scenarios, such as a medic using the Availability Heuristic (based on Exemplar 3 from Coleman et al., 2025).
- Pending Reasoning Enhancements: Introduced cardinality axioms (e.g., every
HeuristicGuidedProcessrequires aguided_byrelation) and SHACL constraints forMedicalTriageDecisionProcess. - Bug Fix: Corrected syntax error in
AvailabilityHeuristicrestriction (changed toBFO_0000056).
- New Object Properties:
- Previous Version: http://www.heuristicontology.com/2025-05-26/HeuristicOntology; http://www.heuristicontology.com/2025-05-12/HeuristicOntology
- Protégé: Version 5.5 or later for ontology editing.
- OWL Reasoner: HermiT or Pellet for consistency checking.
- RDF Libraries: Apache Jena or RDFLib for programmatic access.
- SHACL Validator: TopBraid SHACL API or similar for constraint validation.
- Download
HeuristicOntology - 2025-08-16-Complete.ttlfrom the Releases page. - Open in Protégé via
File → Open, ensuring Turtle format is selected. - Run a reasoner (e.g., HermiT) to verify consistency.
- Validate instances against SHACL constraints using a SHACL validator.
Retrieve all medical triage decisions and their guiding heuristics:
PREFIX : <http://www.heuristicontology.com/HO0000000272/>
SELECT ?decision ?heuristic ?output
WHERE {
?decision a :MedicalTriageDecisionProcess ;
:guided_by ?heuristic ;
:has_output ?output .
}Parse the ontology using RDFLib in Python:
from rdflib import Graph
g = Graph()
g.parse("HeuristicOntology - 2025-05-26-Complete.ttl", format="turtle")
print("Ontology loaded successfully.")The ontology imports the following:
- Basic Formal Ontology (BFO 2020): http://purl.obolibrary.org/obo/bfo/2020/bfo-core.owl
- Common Core Ontologies (CCO 2.0):
- AgentOntology: https://www.commoncoreontologies.org/2024-11-05/AgentOntology
- CommonCoreOntologiesMerged: https://www.commoncoreontologies.org/2024-11-06/CommonCoreOntologiesMerged
- ExtendedRelationOntology: https://www.commoncoreontologies.org/2024-11-06/ExtendedRelationOntology
- GeospatialOntology: https://www.commoncoreontologies.org/2024-11-06/GeospatialOntology
- InformationEntityOntology: https://www.commoncoreontologies.org/2024-11-06/InformationEntityOntology
- TimeOntology: https://www.commoncoreontologies.org/2024-11-06/TimeOntology
Ensure these ontologies are accessible (e.g., via internet or local copies) when loading the Heuristic Ontology.
- Medical Triage: Model heuristic-guided decisions in mass-casualty scenarios, such as applying the START protocol or detecting biases like the Availability Heuristic. For example, the ontology can represent a medic prioritizing a patient based on recalled similar cases.
- Cognitive Science: Analyze bounded rationality and heuristic strategies in decision-making under uncertainty, supporting research in behavioral economics and psychology.
- AI Systems: Integrate heuristics into decision-support systems, cognitive architectures, or explainable AI frameworks to enhance human-aligned decision-making.
- Knowledge Graphs: Link heuristic-guided processes to clinical data for auditing triage decisions or integrating with electronic health records.
The following instance models a medic using the Availability Heuristic to prioritize a patient (from Exemplar 3, Coleman et al., 2025):
:AvailabilityHeuristicDecision123 rdf:type :AvailabilityBasedDecisionProcess , :MedicalTriageDecisionProcess ;
:guided_by :AvailabilityHeuristic03 ;
:has_input :PreviousSimilarPatientCase456 ;
:has_participant :MedicResponder456 ;
:has_participant :Patient123 ;
:has_output :HighPriorityTriageAssignment789 ;
:responseTime "PT5S"^^xsd:duration ;
:cognitiveLoad "3"^^xsd:integer .heuristic-ontology/
├── HeuristicOntology - 2025-05-26-Complete.ttl
├── README.md
├── LICENSE
├── CHANGELOG.md
├── CONTRIBUTING.md
├── docs/
│ ├── examples.md
│ ├── usage-guide.md
├── examples/
│ ├── triage-scenario-1.ttl
│ ├── sparql-queries.rq
├── scripts/
│ ├── validate.py
│ ├── query.py
└── .gitignore
We welcome contributions to refine the heuristic typology, add domain-specific heuristics (e.g., clinical diagnostics), or enhance reasoning capabilities. To contribute:
- Fork the repository.
- Create a branch (
git checkout -b feature/new-heuristic). - Commit changes (
git commit -m "Added ClinicalDiagnosticHeuristic class"). - Submit a pull request with a clear description and rationale.
- Ensure changes align with BFO 2020 and CCO 2.0 standards.
- Validate changes with Protégé and a SHACL validator.
Report issues (e.g., syntax errors, missing classes) via the Issues tab.
See CONTRIBUTING.md for detailed guidelines.
This ontology is licensed under the Creative Commons Attribution 4.0 International License (CC BY 4.0). See LICENSE for details.
- Contributors:
- Timothy W. Coleman
- John H Bittner II
- Affiliations: SUNY University at Buffalo, National Center for Ontological Research
- Reference: Coleman, T. W., & Bittner, J. H. (2025). Formalizing Heuristics: Cognitive Strategies for Decisions Under Constraint.
- Contact: Open an issue or email the contributors for inquiries.
- Issue Tracker: https://github.com/TimothyWColeman/heuristic-ontology/issues
The Heuristic Ontology adheres to FAIR (Findable, Accessible, Interoperable, Reusable) principles:
- Findable: Uses a persistent IRI and is documented for discovery.
- Accessible: Freely available on GitHub under CC BY 4.0.
- Interoperable: Aligned with BFO 2020 and CCO 2.0, using Turtle format.
- Reusable: Includes examples, scripts, and clear documentation.
This work builds on foundational research in cognitive science and ontology engineering, with gratitude to the BFO and CCO communities for their standards and support.