How eitiyamamoto/full-triple-matcher works
Full-Triple-Matcher is a research project for integrating general-purpose and domain-specific knowledge graphs through entity and schema alignment. The documented workflow computes functionality and inverse functionality, runs the early-stop FTM matcher, and evaluates matches; triple matching can also be run separately from entity pairs. The README also describes experimental evaluation and broader alignment methods. Source excerpts cover only graph similarity helpers, so most workflow relationships below are based on explicit README statements, not inspected notebook internals.
Main entry points: FTM - early stop.ipynb Early-Stop FTM · results Matcher Results · FTM - Calculate inverse functionality.ipynb Functionality Metrics · Analyze paris and ftm matches.ipynb Match Metrics
Components 13
Entity Matching
- Functionality Metrics
codes/entity-matching/FTM - Calculate inverse functionality.ipynb - Early-Stop FTM
codes/entity-matching/FTM - early stop.ipynb - Entity and Schema Alignment
codes/entity-matching - Graph Similarity
codes/entity-matching/graph_similarity.py
Evaluation
- Top-N Analysis
codes/entity-matching/Analyzing top@n.ipynb - Match Metrics
codes/entity-matching/Analyze paris and ftm matches.ipynb
Triple Matching
- Triple Similarity
codes/triple-matching/Calculate triple similarity.ipynb - Triple Gold Standards
codes/triple-matching/Creating gold standard.ipynb
Data and Results
- KGs and Seeds
dataset/entity-matching - Entity Gold Standards
dataset/entity-matching - Triple Data
codes/triple-matching - Matcher Results
results
Other components
- Researcher
Connections 17
- Researcher → Functionality Metrics: runs
README.md - Functionality Metrics → Early-Stop FTM: precomputes metrics
README.md - Researcher → Early-Stop FTM: runs
README.md - KGs and Seeds → Functionality Metrics: provides KGs
README.md - KGs and Seeds → Early-Stop FTM: provides KGs
README.md - Early-Stop FTM → Matcher Results: writes matches
README.md - Researcher → Top-N Analysis: runs
README.md - Researcher → Match Metrics: runs
README.md - Matcher Results → Top-N Analysis: evaluates matches
README.md - Matcher Results → Match Metrics: evaluates matches
README.md - Entity Gold Standards → Match Metrics: supplies gold standards
README.md - Researcher → Triple Similarity: runs optionally
README.md - Matcher Results → Triple Similarity: provides entity pairs
README.md - Triple Data → Triple Similarity: provides triples
- Researcher → Triple Gold Standards: creates standards
- Graph Similarity → Early-Stop FTM: supports comparison
- Researcher → Entity and Schema Alignment: uses alignment methods
README.md