Table 14 Performance comparison of the proposed model with state-of-the-art.

From: An efficient learning based approach for automatic record deduplication with benchmark datasets

Dataset

F1- score

MDedup48

Hierarchical NN49

Proposed model (with all extensions)

Amazon-google

0.56

0.68

0.84

Amazon-Walmart dataset

0.49

0.58

0.879

dblp-scholar

0.83

0.90

0.99