{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001375932,0.0001465987,0.0002619991,0.00007251336,0.000127985,0.0001687586,0.001692415,0.0001076229,0.00002323084],"category_scores_gemma":[0.001317032,0.00009297847,0.0001363469,0.0004374479,0.00009780462,0.0004250038,0.00007254087,0.0004327817,0.000002794761],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006093048,"about_ca_system_score_gemma":0.0001574119,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001439777,"about_ca_topic_score_gemma":0.001674859,"domain_scores_codex":[0.998243,0.0003202494,0.0004847043,0.0002731844,0.000468711,0.0002101457],"domain_scores_gemma":[0.9967984,0.00136054,0.0007899746,0.0005886846,0.000433345,0.00002903786],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0001027995,0.0003502447,0.08872494,0.0000466034,0.0001092524,7.42493e-7,0.009573753,0.00003062443,0.2378432,0.2384486,0.0002277562,0.4245416],"study_design_scores_gemma":[0.006779262,0.0002823292,0.2169532,0.0001607933,0.0001421356,0.000005804708,0.0006891557,0.1757808,0.117651,0.4681767,0.01203432,0.00134459],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6821426,0.0009253173,0.2645887,0.04725928,0.001460431,0.001983872,0.0003610358,0.000370277,0.0009085626],"genre_scores_gemma":[0.9496322,0.000009165571,0.04966901,0.0004523382,0.00005172357,0.00006400955,0.00002633804,0.000008069909,0.00008719254],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.423197,"threshold_uncertainty_score":0.3791552,"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."}}