ontology extension

ontology extension extends ontologies by identifying words and phrases denoting ontology classes using lexical features and context from Europe PMC full-text articles to support class discovery and subclass assignment.

Key Features:

• Lexical–context integration: Integrates lexical features with context information from Europe PMC full-text articles to identify mentions of ontology classes.
• Lexicon expansion: Expands lexicons using known labels and synonyms as a basis for exploring additional class labels and potential super-classes.
• Word embeddings: Uses word embeddings to capture contextual nuances and enhance recognition of lexical variants related to ontology classes.
• Machine learning and neural classification: Applies machine learning algorithms, including an artificial neural network, to classify candidate terms corresponding to existing ontology classes not yet included.
• Automated ontology reasoning: Enables automated reasoning over ontologies to improve accuracy and relevance in class identification.
• Subclass categorization: Determines whether a class should be categorized as a subclass under high-level ontology classes to support ontology extension and quality control.

Scientific Applications:

• New label discovery: Aids identification of new labels for ontology classes from literature.
• Disease ontology curation: Identifies disease-related terms and distinguishes disease types within the Human Disease Ontology.
• Ontology extension and quality control: Supports extending ontologies and validating subclass relationships for quality control.

Methodology:

Integrates lexical features with context from Europe PMC full-text articles, expands lexicons using known labels and synonyms, uses word embeddings to capture contextual nuances, applies machine learning algorithms including an artificial neural network for classification, and employs automated reasoning to determine subclass relationships.