Scientific Evidence and Provenance Information Ontology (SEPIO)

The Scientific Evidence and Provenance Information Ontology (SEPIO) is an ontology for representing scientific evidence and provenance information, especially in relation to scientific claims [1] [2]. SEPIO provides a structured vocabulary for describing claims, evidence lines, information items, methods, tools, agents, and the provenance relationships involved in the creation and evaluation of scientific assertions [2].

The ontology employs a class-based modeling approach, defining classes for different types of scientific evidence, provenance, claims, evidence lines, and related information objects, along with properties to describe their relationships and interactions [1] [2]. Hierarchies and relations are used to organize evidence and provenance information, enabling structured retrieval, analysis, and comparison of scientific claims [2]. SEPIO supports the integration of evidence and provenance metadata from various sources, promoting interoperability and data-driven research in scientific evidence representation [1] [2].

Typical applications of SEPIO include scientific claim annotation, evidence modeling, provenance tracking, data integration, curation, knowledge discovery, and manual or computational evaluation of scientific claims [2]. By providing a standardized vocabulary and framework, SEPIO enhances interoperability and supports richer analysis of scientific evidence and provenance across research data platforms [1] [2].

Example Usage: Annotate a scientific study with SEPIO terms to specify a scientific claim, the evidence lines supporting or evaluating it, the information items used as evidence, and the methods, tools, and agents involved in producing that evidence. This enables semantic search, provenance tracking, claim evaluation, and integration with scientific evidence platforms [1] [2].

Metrics & Statistics

Graph Statistics

Total Nodes

1262

Total Edges

2385

Root Nodes

72

Leaf Nodes

781
Knowledge Coverage Statistics

Classes

129

Individuals

21

Properties

117
Hierarchical Metrics

Maximum Depth

14

Minimum Depth

0

Average Depth

3.17

Depth Variance

8.36
Breadth Metrics

Maximum Breadth

170

Minimum Breadth

1

Average Breadth

40.60

Breadth Variance

2186.24
LLMs4OL Dataset Statistics

Term Types

21

Taxonomic Relations

141

Non-taxonomic Relations

6

Average Terms per Type

4.20

Usage Example

Use the following code to import this ontology programmatically:

from ontolearner.ontology import SEPIO

ontology = SEPIO()
ontology.load("path/to/SEPIO-ontology.owl")

# Extract datasets
data = ontology.extract()

# Access specific relations
term_types = data.term_typings
taxonomic_relations = data.type_taxonomies
non_taxonomic_relations = data.type_non_taxonomic_relations

References