{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.004732293,0.001384967,0.002258646,0.01319113,0.00199648,0.0034039,0.002915204,0.002108155,0.002847876],"category_scores_gemma":[0.02882311,0.0006221484,0.0009333476,0.006698724,0.0007302444,0.004004504,0.002540394,0.00124018,0.005530773],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007902823,"about_ca_system_score_gemma":0.001514868,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002885959,"about_ca_topic_score_gemma":0.005068754,"domain_scores_codex":[0.9943955,0.0007335879,0.0006459375,0.001624031,0.002296928,0.0003039555],"domain_scores_gemma":[0.9669638,0.01132428,0.004385382,0.00721957,0.009284232,0.0008226985],"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":[0.001075921,0.0006973902,0.03730682,0.001615966,0.0004660214,0.001995494,0.00189798,0.004866892,0.04164894,0.002707035,0.07853556,0.827186],"study_design_scores_gemma":[0.0002857282,0.001143778,0.07819527,0.0004559723,0.000781611,0.01438212,0.003442245,0.3695964,0.3361073,0.01996044,0.1752082,0.0004409377],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.4631116,0.00979627,0.4269419,0.001969106,0.001939873,0.001251881,0.01849396,0.05526827,0.02122715],"genre_scores_gemma":[0.4751827,0.001816236,0.4780611,0.0006617074,0.001147589,0.0003616406,0.02370477,0.001508905,0.01755542],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01319113,"threshold_uncertainty_score":0.0250271,"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."}}