{"id":"W3172252280","doi":"","title":"WeBiText: Multilingual Concordancer Built from Public High Quality Web Content","year":2010,"lang":"en","type":"article","venue":"Conference of the Association for Machine Translation in the Americas","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Computer science; Machine translation; Quality (philosophy); Web page; Word (group theory); Order (exchange); Translation (biology); Envelope (radar); World Wide Web; Information retrieval; Natural language processing; Artificial intelligence; Telecommunications; Mathematics","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"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":[0.00423794,0.001755122,0.001820338,0.006222053,0.001611937,0.002856015,0.002158354,0.001578181,0.02158032],"category_scores_gemma":[0.02284973,0.001154562,0.001094645,0.005461341,0.0007647556,0.00577141,0.003487099,0.00159495,0.01805545],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001175955,"about_ca_system_score_gemma":0.002307243,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0076834,"about_ca_topic_score_gemma":0.007869518,"domain_scores_codex":[0.9933335,0.001305607,0.0007411118,0.001254243,0.003032389,0.0003330972],"domain_scores_gemma":[0.9836701,0.006161117,0.0009247561,0.004136521,0.004455701,0.0006517626],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.003676167,0.0008521362,0.01714535,0.002819852,0.0004612999,0.004075143,0.003690054,0.004730786,0.06420143,0.008471955,0.2434096,0.6464663],"study_design_scores_gemma":[0.000836063,0.00110156,0.03693625,0.0006061557,0.0003914737,0.006175858,0.002368617,0.1318188,0.2605506,0.01266543,0.5455346,0.001014506],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"software","genre_gemma":"software","genre_scores_codex":[0.09760901,0.001129905,0.2807955,0.0006012387,0.0005482016,0.001730859,0.03471414,0.5487015,0.03416968],"genre_scores_gemma":[0.2978279,0.0006368231,0.4535072,0.0006495521,0.0002988733,0.001239726,0.158267,0.05435299,0.03321998],"genre_candidate":"software","genre_consensus":"software","teacher_disagreement_score":0.02158032,"threshold_uncertainty_score":0.07219332,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08095440750312109,"score_gpt":0.3455252792081267,"score_spread":0.2645708717050056,"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."}}