{"id":"W2095998031","doi":"10.1109/hicss.2007.98","title":"Automatic Web Page Categorization using Principal Component Analysis","year":2007,"lang":"en","type":"article","venue":"","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University","funders":"","keywords":"Computer science; Information retrieval; Web page; World Wide Web; Categorization; Static web page; Web search engine; Web query classification; Web search query; Task (project management); Web navigation; Search engine; Artificial intelligence; Engineering","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":[0.001487565,0.001235912,0.001458816,0.01156312,0.0009060558,0.002151213,0.0009480223,0.001037918,0.002952721],"category_scores_gemma":[0.005886842,0.0003704094,0.001045301,0.006652709,0.0004575567,0.001823268,0.0008464379,0.001067537,0.003962186],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005605259,"about_ca_system_score_gemma":0.001148236,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005889399,"about_ca_topic_score_gemma":0.003971085,"domain_scores_codex":[0.9978055,0.0005911747,0.0001657277,0.0004223959,0.0007579679,0.0002573113],"domain_scores_gemma":[0.9963765,0.001283064,0.0002491367,0.0003899641,0.001604677,0.00009669042],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0003430061,0.0004198721,0.006697768,0.0002682935,0.0001180619,0.0001480814,0.0001834419,0.007468277,0.02933198,0.002491935,0.01330027,0.9392291],"study_design_scores_gemma":[0.00007522674,0.0002783872,0.03965211,0.00008036083,0.0002091755,0.000626759,0.0004191346,0.8778141,0.04635718,0.01805068,0.01625223,0.0001845677],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1150714,0.001646015,0.8555176,0.0004091069,0.0002022239,0.0008142796,0.001952376,0.02019374,0.004193178],"genre_scores_gemma":[0.3418174,0.0009137819,0.6436865,0.00009282494,0.0001532558,0.0005420665,0.007475111,0.0005666577,0.004752324],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01156312,"threshold_uncertainty_score":0.01171023,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03073576962859396,"score_gpt":0.2876970805385032,"score_spread":0.2569613109099092,"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."}}