CardioBERTa.nl_P_enriched / training_data_stats.json
fernandogd's picture
Release CardioBERTa-based clinical concept disambiguation model
37bad45 verified
Raw History Blame Contribute Delete
971 Bytes
{
"triplets": 1677259,
"malformed_rows": 0,
"unique_triplets": 1667152,
"duplicate_triplets": 10107,
"unique_cuis": 476525,
"unique_anchors": 409633,
"unique_positives": 405646,
"unique_terms": 543012,
"unique_anchor_positive_pairs": 1657274,
"ambiguous_terms_across_cuis": 198416,
"self_pairs_exact": 0,
"self_pairs_normalized": 1820,
"cuis_with_multiple_terms": 476487,
"terms_per_cui_mean": 3.9814070615392687,
"terms_per_cui_median": 3,
"terms_per_cui_p95": 9.0,
"terms_per_cui_max": 36,
"triplets_per_cui_mean": 3.5197712606893656,
"triplets_per_cui_median": 2,
"triplets_per_cui_p95": 10.0,
"triplets_per_cui_max": 35,
"anchor_words_mean": 4.239899741184874,
"positive_words_mean": 4.57059345038542,
"cuis_common_with_synonyms": 79985,
"cuis_added_vs_synonyms": 396540,
"cuis_missing_vs_synonyms": 0,
"terms_common_with_synonyms": 156019,
"terms_added_vs_synonyms": 386993,
"terms_missing_vs_synonyms": 0
}