CardioBERTa.it_GP_enriched / training_data_stats.json
fernandogd's picture
Release CardioBERTa-based clinical concept disambiguation model
176e45b verified
Raw
History Blame Contribute Delete
973 Bytes
{
"triplets": 4714271,
"malformed_rows": 0,
"unique_triplets": 4712493,
"duplicate_triplets": 1778,
"unique_cuis": 476970,
"unique_anchors": 450416,
"unique_positives": 461831,
"unique_terms": 529487,
"unique_anchor_positive_pairs": 4654414,
"ambiguous_terms_across_cuis": 193240,
"self_pairs_exact": 0,
"self_pairs_normalized": 188,
"cuis_with_multiple_terms": 476969,
"terms_per_cui_mean": 9.828039499339582,
"terms_per_cui_median": 7.0,
"terms_per_cui_p95": 27.0,
"terms_per_cui_max": 36,
"triplets_per_cui_mean": 9.883789336855568,
"triplets_per_cui_median": 7.0,
"triplets_per_cui_p95": 31.0,
"triplets_per_cui_max": 35,
"anchor_words_mean": 4.484596451922259,
"positive_words_mean": 4.685869564986824,
"cuis_common_with_synonyms": 69631,
"cuis_added_vs_synonyms": 407339,
"cuis_missing_vs_synonyms": 0,
"terms_common_with_synonyms": 136720,
"terms_added_vs_synonyms": 392767,
"terms_missing_vs_synonyms": 0
}