The basic vocabulary (glossary.txt) is meant as dynamic input to generic Large Language Models to give them up-to-date context on Ethereum terms and definitions.
The generated ontology (ethereum_skos.ttl) is a more full-featured variant useful in orchestrated systems using a graph database.
This repository combines the following sources into a single clean text file in term : definition format. That file (glossary.txt) contains duplicated entries from the sources.
Sources:
- The Ethereum Foundation's Ethereum Glossary
- Consensys' A Blockchain Glossary for Beginners
- The Ethereum Foundation's active Ethereum Improvement Proposals (EIPs)
ChatGPT 4o (Enterprise edition) was used to extract newly-created terms from the EIPs and add them sequentially to the text file until they are all represented.
A python script (mk_skos.py) is used to generate a de-duplicated SKOS ontology. Where multiple definitions exist for the same term, both are kept both using different predicates (skos:definition for the primary and rdfs:comment for the secondary). EIPs are linked via skos:broader where they are found.
The SKOS file parses cleanly into the Protege ontology editor.
Please summarise the newly-introduced terms defined in these documents. Add the EIP number without any hyperlinks to the end of the definition, enclosed in parentheses. Your output format should be "term : definition (EIP number)"
Use succinct terms. For example, when a new opcode is introduced, use the opcode name as the term.