Tree Edit Distance

The conceptual precursor to tree-based measures. A mathematical function that quantifies the textual likeness of two sequences.

• Levenshtein Distance: The minimum number of single-character edits (insertions, deletions, substitutions) to change one word into another.
• Damerau-Levenshtein: Adds transposition of adjacent characters to the operation set.
• Jaccard Coefficient: Measures overlap between sets of n-grams.

These lexical matchers serve as primary filters in ontology alignment pipelines before structural comparison occurs.

The computational process of determining correspondences between heterogeneous ontologies to achieve semantic interoperability.

• Also known as ontology matching.
• Produces an alignment: a set of mappings between semantically equivalent entities.
• Tree Edit Distance is a critical structural matcher within alignment systems.
• Addresses semantic heterogeneity—the divergence in meaning across schemas.

Modern systems combine multiple matchers (lexical, structural, extensional) to generate a unified similarity matrix.

A neural ontology matching technique that encodes structural neighborhood into embedding vectors.

• Uses Graph Convolutional Networks (GCNs) to aggregate features from neighboring nodes.
• Entities with similar local graph topologies map to proximate points in a unified vector space.
• Geometric proximity indicates semantic equivalence.
• Contrasts with classical tree edit distance by learning representations rather than computing explicit edit scripts.

Particularly effective when label heterogeneity is high but structural patterns remain conserved.

A quantitative evaluation metric assessing the logical consistency of an alignment.

• Checks if the merged ontology introduces unsatisfiable classes or disjointness violations.
• A structurally minimal tree edit distance does not guarantee logical coherence.
• Alignments must satisfy the Conservativity Principle: no new subsumption relationships should emerge between named classes.

Tools like LogMap use logic-based reasoning to repair incoherent mappings post-alignment, ensuring the merged taxonomy remains logically sound.

An algorithmic solution for finding a stable one-to-one mapping between two sets based on mutual preference scores.

• Applied to ontology matching to optimize global alignment cardinality.
• Each entity ranks potential matches; the algorithm resolves conflicts to prevent blocking pairs.
• Ensures no two entities would prefer each other over their current assigned match.
• Contrasts with greedy nearest-neighbor matching, which can produce suboptimal global alignments.

Used when a strict 1:1 correspondence is required between taxonomy nodes.

A mathematical method for deriving a concept lattice from a formal context of objects and attributes.

• Used for bottom-up taxonomy induction and ontology restructuring.
• Identifies natural groupings of entities based on shared attribute sets.
• The resulting lattice reveals hierarchical relationships that may not be explicit in the original tree structure.
• Complements tree edit distance by providing an alternative structural representation for comparison.

Particularly useful when aligning flat thesauri that lack explicit hierarchical organization.