{"id":"W2289946673","doi":"10.1145/2808797.2809373","title":"Sentiment Crawling","year":2015,"lang":"en","type":"article","venue":"","topic":"Spam and Phishing Detection","field":"Computer Science","cited_by":25,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University","funders":"","keywords":"Web crawler; Crawling; Computer science; Web page; Focused crawler; Class (philosophy); The Internet; World Wide Web; Software; Tree (set theory); Process (computing); Information retrieval; Decision tree; Data mining; Artificial intelligence; Static web page; Web navigation","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":[0.002782339,0.001065794,0.001130729,0.00541456,0.001511461,0.001823148,0.0008367084,0.0006730462,0.003797012],"category_scores_gemma":[0.01061129,0.0005495598,0.001214592,0.003387196,0.0003244572,0.00155256,0.0009582749,0.0009526056,0.004623084],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007253198,"about_ca_system_score_gemma":0.001395758,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003056798,"about_ca_topic_score_gemma":0.006492562,"domain_scores_codex":[0.9970508,0.0007244451,0.0003647832,0.0005581672,0.001106262,0.0001956738],"domain_scores_gemma":[0.992187,0.001927994,0.0004405085,0.001060102,0.004202616,0.0001817937],"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":[0.0003460422,0.0005919031,0.03144912,0.001434231,0.0002859015,0.0005430157,0.001781521,0.003859142,0.05557304,0.004844755,0.0913309,0.8079605],"study_design_scores_gemma":[0.0002100013,0.001023109,0.1131368,0.0006412242,0.0006152038,0.002328915,0.002588709,0.2920426,0.1917085,0.01591014,0.3794948,0.0002999511],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3452237,0.001963202,0.4816565,0.001930093,0.001602846,0.009974248,0.0442225,0.03679734,0.07662959],"genre_scores_gemma":[0.3258266,0.001533308,0.573642,0.001388101,0.0005130888,0.00479023,0.05984155,0.002604028,0.02986112],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.00541456,"threshold_uncertainty_score":0.0147146,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04139787596313377,"score_gpt":0.2505965403220559,"score_spread":0.2091986643589221,"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."}}