Sparkly: A Simple yet Surprisingly Strong TF/IDF Blocker for Entity Matching

Benchmark Model Rank Results
blocking-on-abt-buySparkly k=10#4Candidate Set Size: 10900Recall: 98.1
blocking-on-abt-buySparkly k=50#6Candidate Set Size: 54500Recall: 99.2
blocking-on-amazon-googleSparkly k=10#2Candidate Set Size: 33300Recall: 96.8
blocking-on-amazon-googleSparkly k=50#6Candidate Set Size: 165900Recall: 99.2