{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001150859,0.0006991406,0.000852804,0.01234887,0.001043486,0.002380988,0.0009386565,0.001033074,0.001298294],"category_scores_gemma":[0.006568035,0.0002190858,0.0008652951,0.01104994,0.0006827355,0.001857336,0.001452846,0.0007281302,0.001516531],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006896973,"about_ca_system_score_gemma":0.0009342017,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.009309815,"about_ca_topic_score_gemma":0.01131523,"domain_scores_codex":[0.9979076,0.0004498777,0.0002698115,0.0003558686,0.0007927511,0.0002240464],"domain_scores_gemma":[0.9959636,0.001498919,0.0007527493,0.0005560842,0.0009921698,0.000236453],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00137735,0.001158929,0.3064897,0.0009569103,0.0004707242,0.002579404,0.003116105,0.07609717,0.02248307,0.01359879,0.02345599,0.5482159],"study_design_scores_gemma":[0.00005143122,0.0001920761,0.1573938,0.0001184721,0.0002015766,0.001584251,0.006112346,0.77095,0.01255934,0.0323045,0.01834909,0.0001831948],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.7098331,0.001943607,0.2567517,0.001213983,0.0002523528,0.000410089,0.0172947,0.003666124,0.008634322],"genre_scores_gemma":[0.8126808,0.000623399,0.1606849,0.0001367016,0.0001940642,0.0001847078,0.02337764,0.0001822261,0.001935601],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.01234887,"threshold_uncertainty_score":0.0185113,"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."}}