Entity resolution identifies and merges different representations of the same real-world entity in a dataset or knowledge graph. It removes duplicate nodes so relationships, inferences, and queries remain consistent. section 6.2.2 section 6.3
Common approaches include:
- String matching for spelling variations and naming differences. section 6.2.2
- Clustering algorithms for grouping likely duplicate references. section 6.2.2
- Machine-learning methods that use surrounding context to distinguish duplicates. section 6.2.2
- Domain-specific ontologies or rules, with subject-matter expertise to define suitable decisions. section 6.2.2
Organization name variants
UTI Asset Management Company, UTI Asset Management Company Limited, and UTI Asset Management Company Ltd can represent one organization. Resolution unifies them into a single coherent graph node. section 6.2.2
Figure 6.5 Potential duplicates
Inspect the potential-duplicates figure to compare multiple graph nodes representing the same entity. section 6.2.2
