Datasets:
Formats:
csv
Languages:
English
Size:
1M - 10M
ArXiv:
Tags:
pearl benchmark
phrase embeddings
entity retrieval
entity clustering
fuzzy join
entity matching
License:
[bot] Conversion to Parquet
#1 opened 9 months ago
by
parquet-converter