{"id":"W3175120204","doi":"10.1109/wiiat50758.2020.00087","title":"Analytics of Similar-Sounding Names from the Web with Phonetic Based Clustering","year":2020,"lang":"en","type":"article","venue":"","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":16,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Manitoba","funders":"University of Manitoba","keywords":"Computer science; Cluster analysis; Big data; Task (project management); Web intelligence; Web mining; World Wide Web; Semantic Web; Analytics; Data science; Focus (optics); Depth sounding; Data Web; Information retrieval; The Internet; Web page; Data mining; Artificial intelligence; Web modeling; Geography; Engineering","routes":{"ca_aff":true,"ca_fund":true,"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":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.001099383,0.00008189312,0.0001837418,0.00003861399,0.00007266378,0.0002039253,0.0008052815,0.00001815294,0.001269259],"category_scores_gemma":[0.0008426341,0.00004078688,0.00005057412,0.0005592775,0.00009131505,0.0001413874,0.0002669993,0.00006427043,0.0000705541],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000007864112,"about_ca_system_score_gemma":0.0000367126,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001114029,"about_ca_topic_score_gemma":0.0005683009,"domain_scores_codex":[0.998239,0.0001295487,0.0003510306,0.0002695866,0.0008910853,0.0001197444],"domain_scores_gemma":[0.9977862,0.001441333,0.00014024,0.0005040979,0.00006336698,0.0000647819],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001078835,0.000424992,0.1120017,0.0001649663,0.0007661948,0.00009026119,0.01937245,0.1927411,0.005185324,0.01828537,0.5613928,0.08849604],"study_design_scores_gemma":[0.0008888294,0.0001725406,0.007276777,0.00004218425,0.00008099421,2.496854e-7,0.01530491,0.8320733,0.00103986,0.002407948,0.1405013,0.0002110841],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1184109,0.00005974846,0.826493,0.04053304,0.0001608754,0.000269085,0.0001372636,0.00006550304,0.01387056],"genre_scores_gemma":[0.9858864,0.000005378841,0.00663373,0.007133336,0.0000396525,0.000001686857,0.000007596318,0.000005087438,0.0002871589],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8674755,"threshold_uncertainty_score":0.9996437,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2544112191944405,"score_gpt":0.3696189409535637,"score_spread":0.1152077217591231,"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."}}