{"id":"W2333422287","doi":"","title":"OSD: An Online Web Spam Detection System","year":2009,"lang":"en","type":"article","venue":"","topic":"Spam and Phishing Detection","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University","funders":"","keywords":"Spambot; Computer science; Forum spam; Spamdexing; Web page; Information retrieval; World Wide Web; Spamming; Relevance (law); Static web page; Web crawler; The Internet; Web navigation; Web search engine","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0001607467,0.00007781359,0.00007511974,0.00008813916,0.0001085254,0.0001521641,0.0003119978,0.00005728524,0.000006749766],"category_scores_gemma":[0.000008678467,0.00006774593,0.0000322656,0.0002961743,0.000005134278,0.0005993078,0.00002118144,0.00008655258,0.00006566648],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005563808,"about_ca_system_score_gemma":0.00001737971,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00007889949,"about_ca_topic_score_gemma":0.0001644991,"domain_scores_codex":[0.999296,0.00004229117,0.0001193451,0.0002454875,0.0001552304,0.0001416445],"domain_scores_gemma":[0.9994569,0.00001160608,0.00003731931,0.0003751161,0.000042312,0.00007679177],"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":[0.00001994284,0.000238208,0.0002198335,0.00001769162,0.000009719453,0.00001994841,0.0003697763,0.0003613878,0.1138104,0.02390778,0.0004705082,0.8605548],"study_design_scores_gemma":[0.0004143122,0.0009845718,0.01860327,0.00002505946,0.000007588005,0.0001672284,0.0001044325,0.9201192,0.05231733,0.00134835,0.005593042,0.0003156674],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2763468,0.00002110325,0.7147437,0.0004122675,0.001011803,0.00008813723,7.061385e-7,0.001338133,0.006037321],"genre_scores_gemma":[0.9867002,0.000001720369,0.01251028,0.000285579,0.0002659115,0.000001366426,0.000001345638,0.000003319618,0.0002302456],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9197578,"threshold_uncertainty_score":0.2762599,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02180911472220613,"score_gpt":0.2320117895543101,"score_spread":0.210202674832104,"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."}}