entity-resolution
Identifies and merges duplicate entities in Knowledge Graph projects. Use after extraction to consolidate duplicate entities, or when users ask about potential duplicates. Supports automatic merging for high-confidence matches and user confirmation for medium-confidence candidates.
Security Assessment
About entity-resolution
The entity-resolution skill is designed to identify and merge duplicate entities within Knowledge Graph projects. This skill helps consolidate duplicate entities that may arise during data extraction, or when users suspect that multiple entries refer to the same entity. It leverages various signals such as string similarity, alias overlap, type matching, and graph context to detect potential duplicates. This ensures that the Knowledge Graph remains clean and consistent, with minimal manual intervention. Whether automatically merging high-confidence matches or prompting users for confirmation on medium-confidence candidates, the skill streamlines the process of maintaining entity integrity in large-scale graph-based projects.
FAQ
How does the entity-resolution skill detect duplicates?
It uses several techniques including string similarity (Jaro-Winkler), alias overlap (Jaccard similarity), type matching, and graph context (shared neighbors).
Can the skill automatically merge entities?
Yes, the skill can automatically merge entities with high confidence (>= 0.9). For medium-confidence matches (0.7-0.9), it will prompt the user for confirmation.
Is there any way to review pending merges?
Yes, you can review pending merges using the `review_pending_merges` tool, which shows candidates awaiting user approval.
What happens when two entities are merged?
The survivor entity keeps its primary label, while the merged entity’s label becomes an alias. All relationships, properties, and aliases are transferred to the survivor.
Does the entity-resolution skill work with any Knowledge Graph platform?
The skill is designed to work within projects using a specific Knowledge Graph platform, and its effectiveness depends on compatibility with the graph's structure.
Install entity-resolution
Quick Setup:
- Copy the skill folder to
.claude/skills/ - Claude will automatically detect and use the skill
Repository
costiash/CognivAgent