{"id":"W4293243715","doi":"10.23889/ijpds.v7i3.2077","title":"Analysing Siamese Neural Network Architectures for Computing Name Similarity.","year":2022,"lang":"en","type":"article","venue":"International Journal for Population Data Science","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Guelph","funders":"","keywords":"Similarity (geometry); Computer science; Classifier (UML); False positive paradox; Artificial intelligence; String metric; Artificial neural network; Random forest; Similitude; False positives and false negatives; Set (abstract data type); Machine learning; Pattern recognition (psychology); Data mining; String searching algorithm; Pattern matching","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["sts","scholarly_communication","open_science"],"consensus_categories":[],"category_scores_codex":[0.01694581,0.0001164302,0.0001904095,0.0006097807,0.002467159,0.001879167,0.006813873,0.00001607516,0.0001753222],"category_scores_gemma":[0.006043157,0.00009992385,0.0001303243,0.0008947331,0.0001402122,0.001229532,0.002741666,0.0002327338,0.000003610577],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001785661,"about_ca_system_score_gemma":0.0001244313,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00008290714,"about_ca_topic_score_gemma":0.00008549764,"domain_scores_codex":[0.9944961,0.0001951135,0.000896965,0.0006772706,0.003357423,0.0003770957],"domain_scores_gemma":[0.9965632,0.001148295,0.000778234,0.0007244835,0.0006561311,0.0001296652],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000124325,0.00005186547,0.01074011,0.000002104395,0.00003960551,0.000005322761,0.0001978773,0.8000485,0.00003424475,0.01151757,0.02405393,0.1531845],"study_design_scores_gemma":[0.0003622064,0.00005589077,0.0184984,0.00000672401,0.00001597704,0.00005693482,0.0003757065,0.8098908,0.00000319319,0.06038423,0.110222,0.0001279131],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.275097,0.00009608599,0.6992705,0.01161035,0.01076272,0.0006318418,0.002284129,0.00004973104,0.0001976979],"genre_scores_gemma":[0.9677878,0.000001532614,0.02876791,0.001672234,0.0009053072,0.000008665075,0.0006322478,0.000008250809,0.000216044],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.6926908,"threshold_uncertainty_score":0.999157,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2782177487324201,"score_gpt":0.5106037761682547,"score_spread":0.2323860274358346,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}