CardioBERTa.ro_GP_enriched / training_data_stats.json
fernandogd's picture
Release CardioBERTa-based clinical concept disambiguation model
05213cb verified
Raw History Blame Contribute Delete
972 Bytes
{
"triplets": 4734361,
"malformed_rows": 0,
"unique_triplets": 4733225,
"duplicate_triplets": 1136,
"unique_cuis": 476970,
"unique_anchors": 444941,
"unique_positives": 470719,
"unique_terms": 531980,
"unique_anchor_positive_pairs": 4680255,
"ambiguous_terms_across_cuis": 193880,
"self_pairs_exact": 0,
"self_pairs_normalized": 106,
"cuis_with_multiple_terms": 476969,
"terms_per_cui_mean": 9.850546155942721,
"terms_per_cui_median": 7.0,
"terms_per_cui_p95": 28.0,
"terms_per_cui_max": 37,
"triplets_per_cui_mean": 9.92590938633457,
"triplets_per_cui_median": 7.0,
"triplets_per_cui_p95": 31.0,
"triplets_per_cui_max": 35,
"anchor_words_mean": 4.438981100089325,
"positive_words_mean": 4.662949445553476,
"cuis_common_with_synonyms": 70817,
"cuis_added_vs_synonyms": 406153,
"cuis_missing_vs_synonyms": 0,
"terms_common_with_synonyms": 139248,
"terms_added_vs_synonyms": 392732,
"terms_missing_vs_synonyms": 0
}