CardioBERTa.it_P_enriched / training_data_stats.json
fernandogd's picture
Release CardioBERTa-based clinical concept disambiguation model
8113659 verified
Raw History Blame Contribute Delete
967 Bytes
{
"triplets": 1597673,
"malformed_rows": 0,
"unique_triplets": 1596913,
"duplicate_triplets": 760,
"unique_cuis": 476349,
"unique_anchors": 380100,
"unique_positives": 414377,
"unique_terms": 529199,
"unique_anchor_positive_pairs": 1580295,
"ambiguous_terms_across_cuis": 193802,
"self_pairs_exact": 0,
"self_pairs_normalized": 191,
"cuis_with_multiple_terms": 476346,
"terms_per_cui_mean": 3.927168945458057,
"terms_per_cui_median": 3,
"terms_per_cui_p95": 9.0,
"terms_per_cui_max": 36,
"triplets_per_cui_mean": 3.3539967544804337,
"triplets_per_cui_median": 2,
"triplets_per_cui_p95": 9.0,
"triplets_per_cui_max": 35,
"anchor_words_mean": 4.625564805814457,
"positive_words_mean": 4.934238107547665,
"cuis_common_with_synonyms": 69631,
"cuis_added_vs_synonyms": 406718,
"cuis_missing_vs_synonyms": 0,
"terms_common_with_synonyms": 136720,
"terms_added_vs_synonyms": 392479,
"terms_missing_vs_synonyms": 0
}