CardioBERTa.es_GP_enriched / training_data_stats.json
fernandogd's picture
Release CardioBERTa-based clinical concept disambiguation model
a99e6ad verified
Raw History Blame Contribute Delete
971 Bytes
{
"triplets": 4944387,
"malformed_rows": 0,
"unique_triplets": 4885251,
"duplicate_triplets": 59136,
"unique_cuis": 476973,
"unique_anchors": 473363,
"unique_positives": 468287,
"unique_terms": 545308,
"unique_anchor_positive_pairs": 4857894,
"ambiguous_terms_across_cuis": 196573,
"self_pairs_exact": 0,
"self_pairs_normalized": 5550,
"cuis_with_multiple_terms": 476972,
"terms_per_cui_mean": 9.90767192272938,
"terms_per_cui_median": 7,
"terms_per_cui_p95": 28.0,
"terms_per_cui_max": 36,
"triplets_per_cui_mean": 10.366177959758728,
"triplets_per_cui_median": 7,
"triplets_per_cui_p95": 33.0,
"triplets_per_cui_max": 35,
"anchor_words_mean": 5.052073189254806,
"positive_words_mean": 5.347384215677292,
"cuis_common_with_synonyms": 83752,
"cuis_added_vs_synonyms": 393221,
"cuis_missing_vs_synonyms": 0,
"terms_common_with_synonyms": 159688,
"terms_added_vs_synonyms": 385620,
"terms_missing_vs_synonyms": 0
}