Navigating Corporate Language: Developing a Graph-Based Glossary
summary
Builds a Neo4j glossary by extracting keywords and phrases from documents, turning paragraphs into text nodes, terms into term nodes, and links into relationships, then uses a simple rule (must/shall/will requirement filter) to flag requirements. Demo with two docs (EOPS vs maintenance manual) shows adding candidates (example abnormality), moving chosen terms into the glossary, and finding overlapping critical terms like engine and aircraft. Result is searchable, versionable graph that surfaces isolated glossary islands and shared, requirement-bearing terms.
AI-written summary · about
chapters (6)
- 0:00Introduction and motivation
- 3:19Why corporate vocabulary matters
- 10:45Challenges of traditional glossaries
- 13:56How graph databases help
- 18:13Demo: Building a graph glossary
- 27:05Summary and benefits
