| Amount of data | Redundancy detection |
Unimproved FCM redundancy detection | 150 records (50% semantic similarity) | 10% |
Improved FCM redundancy detection based on ontology | 150 records (50% semantic similarity) | 35% |