AI Glossary
The complete dictionary of Artificial Intelligence
Relation Extraction
Automatic extraction of semantic relations between entities from unstructured texts to enrich and extend existing knowledge graphs.
Triple Classification
Classification task aimed at determining whether a triple (subject, predicate, object) is valid in a knowledge graph, essential for KG completion and validation.
Ontology Integration
Harmonious fusion of multiple ontologies or knowledge schemas to create a unified and coherent representation of concepts and their relationships.
TransE Algorithm
Graph embedding model representing relations as vector translations between entities, where h + r ≈ t for a valid triple (h, r, t).
Multi-hop Reasoning
Reasoning process requiring multiple steps of navigation through a knowledge graph to establish indirect relationships between entities.
Structural Embedding
Vector representation preserving the topology and structure of the original graph, capturing relational patterns and implicit hierarchies between entities.
Relational Reasoning
Ability to infer and manipulate complex relationships between multiple entities, essential for reasoning in Few-Shot and Zero-Shot Learning.
Graph-based Representation Learning
Machine learning paradigm aimed at automatically discovering optimal vector representations for nodes, edges, and subgraphs in relational data structures.