CardioBERTa.es_P_enriched / training_data_stats.json
fernandogd's picture
Release CardioBERTa-based clinical concept disambiguation model
feafd30 verified
Raw History Blame Contribute Delete
975 Bytes
{
"triplets": 1696740,
"malformed_rows": 0,
"unique_triplets": 1674026,
"duplicate_triplets": 22714,
"unique_cuis": 476638,
"unique_anchors": 399204,
"unique_positives": 418108,
"unique_terms": 545019,
"unique_anchor_positive_pairs": 1665125,
"ambiguous_terms_across_cuis": 197337,
"self_pairs_exact": 0,
"self_pairs_normalized": 5750,
"cuis_with_multiple_terms": 476628,
"terms_per_cui_mean": 3.9709674847578245,
"terms_per_cui_median": 3.0,
"terms_per_cui_p95": 9.0,
"terms_per_cui_max": 36,
"triplets_per_cui_mean": 3.559808491979238,
"triplets_per_cui_median": 2.0,
"triplets_per_cui_p95": 10.0,
"triplets_per_cui_max": 35,
"anchor_words_mean": 5.186215919940592,
"positive_words_mean": 5.628029043931304,
"cuis_common_with_synonyms": 83752,
"cuis_added_vs_synonyms": 392886,
"cuis_missing_vs_synonyms": 0,
"terms_common_with_synonyms": 159688,
"terms_added_vs_synonyms": 385331,
"terms_missing_vs_synonyms": 0
}