Description:
Extend the current knowledge graph by introducing multiple semantic relationship types beyond “SIMILAR_TO.” Based on the previously identified relationship categories (e.g., requires, enables, prerequisite_for, used_for, applies_to), design and implement a classification mechanism to assign appropriate relationship types between ONET and ESCO concepts.
This task includes:
- Defining clear semantic rules or prompts for each relationship type
- Applying classification (rule-based or LLM-assisted) to existing matched pairs
- Updating edge schema to store relationship types
- Evaluating how multi-relational edges improve graph interpretability and usefulness
The goal is to move from a similarity-based graph to a semantically rich knowledge graph
Description:
Extend the current knowledge graph by introducing multiple semantic relationship types beyond “SIMILAR_TO.” Based on the previously identified relationship categories (e.g., requires, enables, prerequisite_for, used_for, applies_to), design and implement a classification mechanism to assign appropriate relationship types between ONET and ESCO concepts.
This task includes:
The goal is to move from a similarity-based graph to a semantically rich knowledge graph