{"id":"W1882783948","doi":"10.1007/978-3-642-32584-7_30","title":"Classifying Websites into Non-topical Categories","year":2012,"lang":"en","type":"book-chapter","venue":"Lecture notes in computer science","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Alberta","funders":"","keywords":"Computer science; Bigram; Classifier (UML); Information retrieval; Web page; Support vector machine; Set (abstract data type); Artificial intelligence; World Wide Web; Trigram","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":[0.0005723897,0.0006951165,0.0005855349,0.01277723,0.0009250049,0.002951535,0.0006127569,0.0006547003,0.003702189],"category_scores_gemma":[0.002764258,0.0001698089,0.0007827131,0.008338794,0.0004270835,0.002518452,0.0008213472,0.0006831465,0.004754498],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005421661,"about_ca_system_score_gemma":0.0006541652,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001677669,"about_ca_topic_score_gemma":0.0038916,"domain_scores_codex":[0.9989859,0.0001291928,0.00008968969,0.0001912561,0.0004568204,0.0001470784],"domain_scores_gemma":[0.9976308,0.0008015854,0.0002148613,0.0002674783,0.0008842931,0.0002007895],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0002783551,0.0005484176,0.08574121,0.001247845,0.0001680509,0.0003212437,0.001245729,0.001216716,0.03581987,0.008238143,0.03573607,0.8294384],"study_design_scores_gemma":[0.00008523282,0.001091521,0.4213513,0.001498122,0.001298195,0.005396163,0.01189553,0.1678116,0.0861161,0.07510768,0.2281294,0.0002192002],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6757804,0.01061476,0.1744359,0.001176567,0.0007441175,0.0009885762,0.01453028,0.004981613,0.1167478],"genre_scores_gemma":[0.8425495,0.004041582,0.09431517,0.0002580176,0.000569482,0.0003252066,0.02374601,0.0004907468,0.03370427],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01277723,"threshold_uncertainty_score":0.01238507,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02252797099015956,"score_gpt":0.2605807330893218,"score_spread":0.2380527620991623,"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."}}