ONTOLOGY ANALYSIS

Provides a no-SPARQL abstraction layer for querying and comparing entities in an RDF/OWL ontology store. Supports seven analysis modes — list, get, find-related, compare, find-common, find-exclusive, and aggregate — against any named ARC2 triple store. Use this worker when an AI agent or workflow step needs structured ontology introspection without hand-written SPARQL.

When to use

Tagged: aggregate, ai-agent, arc2, compare, entity-analysis, graph-traversal, ontology, owl.

Inputs

Label ID Type Default Required Description
Ontology Store Name ontology_store_name text Name of the ARC2 RDF triple store to analyze; the prefix ‘arc2_ontology_’ is added automatically if absent — e.g. ‘vehicle_model’ resolves to ‘arc2_ontology_vehicle_model’.
Analysis Type analysis_type select The operation to perform: ‘list_entities’, ‘get_entity’, ‘find_related’, ‘compare_entities’, ‘find_common’, ‘find_exclusive’, or ‘aggregate’ — each has specific required companion inputs described below.
Entity Type entity_type text   RDF node type to target, matching an OWL class in the store schema (e.g. ‘SystemModel’, ‘Assembly’); required for all analysis types.
Entity IDs entity_ids text   Comma-separated entity ID(s) to operate on (e.g. ‘SM_001’ or ‘SM_001,SM_002’); for get_entity/find_related up to 5 IDs are accepted, compare_entities requires exactly 2, find_common requires 2+, find_exclusive accepts 1+.
Relationship Path relationship_path text   Comma-separated relationship names defining the traversal path; a single name for direct hops (e.g. ‘hasAssembly’) or multiple names for multi-hop traversal (e.g. ‘hasSimulation,hasResponse’); required for find_related, find_common, find_exclusive, and aggregate.
Direction direction select outgoing   Edge traversal direction for find_related: ‘outgoing’ (source→target, default), ‘incoming’ (target→source), or ‘both’; leave as default unless reverse or bidirectional traversal is needed.
Exclude Entity IDs exclude_entity_ids text   Comma-separated entity IDs to exclude from results; used with find_exclusive to remove specific parent entities from the uniqueness comparison.
Filter Result Properties filter_result_properties text   Comma-separated list of property names to retain in result rows; omit to return all properties.
Results Filter Key results_filter_key text   Legacy single-property filter key; prefer the ‘results_filter’ JSON input for new workflows — used only when results_filter is empty.
Results Filter Value results_filter_value text   Value paired with results_filter_key for legacy single-property filtering; ignored when results_filter JSON is provided.

Outputs

Label ID Type Description
Status status text Machine-readable status string (‘success’ or an error code) indicating whether the analysis operation completed without errors.
Results results dataset Primary analysis output as a tabular dataset; schema varies by analysis_type (entity properties for list/get, traversal results for find_related, counts for aggregate, etc.).
Summary summary text Human-readable prose summary of the analysis findings, suitable for display in a report or agent response.
Result Count count number Integer count of the total number of rows returned in the ‘results’ dataset.
Property Differences property_diff dataset Tabular dataset of property-level differences between the two compared entities; populated only for compare_entities, empty otherwise.
Relationship Differences relationship_diff dataset Tabular dataset of relationship-level differences between the two compared entities; populated only for compare_entities, empty otherwise.
Children Only in First Entity child_only_in_first dataset Dataset of child entity names present exclusively under the first entity; populated only for compare_entities.
Children Only in Second Entity child_only_in_second dataset Dataset of child entity names present exclusively under the second entity; populated only for compare_entities.
Children in Both Entities child_in_both dataset Dataset of child entity names shared by both compared entities; populated only for compare_entities.
Child Similarity Percentage child_similarity_percentage number Percentage of child entity names that overlap between entities (for compare_entities only).
Overall Change Percentage overall_change_percentage number Weighted average of all property change percentages (for compare_entities only).
Nested Properties Differences nested_properties_diff dataset Property-level differences from nested properties objects like Response metrics (for compare_entities only).
Nested Properties Change Percentage nested_properties_change_percentage number Average change percentage for nested properties (for compare_entities only).

Disciplines

  • ai_ml.agents
  • ai_ml.domain_reasoning
  • platform.ontology

Auto-generated from platform schema. Worker id: ontology_analysis. Schema hash: 9657ba91a90b. Hand-curated docs in workerexamples/ override this page when present.