{"id":"W3030133709","doi":"","title":"Online Near-Duplicate Detection of News Articles","year":2020,"lang":"en","type":"article","venue":"NPARC","topic":"Spam and Phishing Detection","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Computer science; Information retrieval; Precision and recall; Context (archaeology); Set (abstract data type); Geotagging; Data mining; Data science; World Wide Web","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00005385586,0.00004947446,0.0000738067,0.00001881578,0.00004713914,0.00004701925,0.0002397585,0.00002894411,0.00001874533],"category_scores_gemma":[0.00005199539,0.0000468613,0.00003597666,0.0003391053,0.00002132303,0.0001768271,0.00006520334,0.00007463215,0.00004167163],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000008057862,"about_ca_system_score_gemma":0.00001495652,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004991187,"about_ca_topic_score_gemma":0.00002244207,"domain_scores_codex":[0.9994789,0.00002369846,0.0001209263,0.0001640293,0.0001200556,0.00009232795],"domain_scores_gemma":[0.9996367,0.00002309161,0.00005274025,0.000187729,0.00003387011,0.00006593904],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.00001034913,0.0000316277,0.0002495293,0.000008272086,0.000004633925,0.000001070593,0.000614982,0.00009519649,0.705612,0.000749693,0.0002352437,0.2923874],"study_design_scores_gemma":[0.0002591557,0.0002752341,0.006605788,0.000006858497,0.000006155051,0.00000527643,0.00002979223,0.3930921,0.5853772,0.007058938,0.007171683,0.0001118867],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6626517,0.00001571686,0.3322914,0.00419253,0.0001233876,0.00005106966,0.000001427259,0.0001428997,0.0005299567],"genre_scores_gemma":[0.9838498,0.000004755907,0.01552501,0.0005134204,0.00008791184,0.00000215738,5.596559e-7,0.000003938401,0.00001239485],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3929969,"threshold_uncertainty_score":0.1910948,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0316744735206537,"score_gpt":0.236338824262735,"score_spread":0.2046643507420813,"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."}}