{"id":"W2096794280","doi":"10.1109/ccece.2011.6030580","title":"Classifying the Arabic web — A pilot study","year":2011,"lang":"en","type":"article","venue":"","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Ottawa","funders":"","keywords":"Computer science; World Wide Web; Arabic; Web crawler; Web page; Globe; Categorization; The Internet; Information retrieval; Artificial intelligence; Linguistics","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001993297,0.0004658303,0.0004051495,0.00133964,0.001147302,0.001097022,0.0003647338,0.0006418427,0.002838944],"category_scores_gemma":[0.01007965,0.0001317547,0.0003680633,0.001144735,0.0005149993,0.001743481,0.0005964189,0.0007178252,0.001451619],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005438881,"about_ca_system_score_gemma":0.0004162694,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006669807,"about_ca_topic_score_gemma":0.006560902,"domain_scores_codex":[0.9990409,0.0004423442,0.0001123114,0.0001221166,0.0002017208,0.00008056662],"domain_scores_gemma":[0.9882141,0.00640486,0.0003889142,0.0008361148,0.003437622,0.0007183758],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.005584292,0.02386433,0.5178056,0.0009325852,0.0002490965,0.003709537,0.03183269,0.006235559,0.03379695,0.001776617,0.01217711,0.3620356],"study_design_scores_gemma":[0.0005775592,0.009255174,0.784659,0.0002161829,0.0003457781,0.002510302,0.06259117,0.05998551,0.03627857,0.002255254,0.04116125,0.0001642877],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9966235,0.00005909163,0.000755977,0.0001209156,0.00001851249,0.0002128852,0.0003429549,0.00003313842,0.001832922],"genre_scores_gemma":[0.9918321,0.0001606784,0.004170998,0.0001118647,0.0000404794,0.0001761808,0.001370343,0.00001678512,0.002120728],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.006669807,"threshold_uncertainty_score":0.01326197,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1525639738533988,"score_gpt":0.272155169388718,"score_spread":0.1195911955353193,"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."}}