{"id":"W1544042013","doi":"10.1002/9781405198431.wbeal0254","title":"Corpus Analysis of the World Wide Web","year":2012,"lang":"en","type":"other","venue":"The Encyclopedia of Applied Linguistics","topic":"Lexicography and Language Studies","field":"Arts and Humanities","cited_by":43,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"World Wide Web; Computer science; Variety (cybernetics); Web standards; Quarter (Canadian coin); Population; Web page; Artificial intelligence; History; Sociology","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":[0.003738694,0.0002824576,0.000376693,0.01244438,0.001325742,0.002623311,0.0006170524,0.0004812812,0.003081419],"category_scores_gemma":[0.02094043,0.000224401,0.0002939983,0.01671449,0.0008756108,0.001804164,0.001594283,0.0007662048,0.0009243623],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007492111,"about_ca_system_score_gemma":0.001266241,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007068827,"about_ca_topic_score_gemma":0.0120751,"domain_scores_codex":[0.9966456,0.001514575,0.0003404955,0.0003690053,0.001043335,0.000086862],"domain_scores_gemma":[0.977236,0.01525103,0.0009345553,0.002339433,0.003995096,0.0002438089],"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":[0.0005722035,0.0005189364,0.05939117,0.003837079,0.0005032482,0.001972983,0.008282823,0.008661252,0.009776495,0.05132959,0.1236299,0.7315243],"study_design_scores_gemma":[0.000129781,0.0001927588,0.1260933,0.00175226,0.0005510709,0.002133812,0.01103786,0.09494299,0.02108042,0.02943427,0.712494,0.000157537],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"other","genre_scores_codex":[0.7164182,0.0210382,0.1054406,0.003096683,0.001601161,0.00114613,0.05978049,0.00284092,0.08863764],"genre_scores_gemma":[0.799343,0.006123303,0.09574214,0.0002208193,0.0003654049,0.001035689,0.08642877,0.0006226737,0.01011831],"genre_candidate":"other","genre_consensus":null,"teacher_disagreement_score":0.01244438,"threshold_uncertainty_score":0.01977229,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01165044583078892,"score_gpt":0.2211673313850941,"score_spread":0.2095168855543052,"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."}}