CardioBERTa.ro_P_enriched / training_data_stats.json
fernandogd's picture
Release CardioBERTa-based clinical concept disambiguation model
1fd1432 verified
Raw History Blame Contribute Delete
969 Bytes
{
"triplets": 1607064,
"malformed_rows": 0,
"unique_triplets": 1606532,
"duplicate_triplets": 532,
"unique_cuis": 476350,
"unique_anchors": 376425,
"unique_positives": 422975,
"unique_terms": 531693,
"unique_anchor_positive_pairs": 1591020,
"ambiguous_terms_across_cuis": 194483,
"self_pairs_exact": 0,
"self_pairs_normalized": 109,
"cuis_with_multiple_terms": 476349,
"terms_per_cui_mean": 3.93920436653721,
"terms_per_cui_median": 3.0,
"terms_per_cui_p95": 9.0,
"terms_per_cui_max": 36,
"triplets_per_cui_mean": 3.3737042090899547,
"triplets_per_cui_median": 2.0,
"triplets_per_cui_p95": 9.0,
"triplets_per_cui_max": 35,
"anchor_words_mean": 4.610658940776472,
"positive_words_mean": 4.86856403976444,
"cuis_common_with_synonyms": 70817,
"cuis_added_vs_synonyms": 405533,
"cuis_missing_vs_synonyms": 0,
"terms_common_with_synonyms": 139248,
"terms_added_vs_synonyms": 392445,
"terms_missing_vs_synonyms": 0
}