{"id":"W123190706","doi":"","title":"Topic Classification of Blog Posts Using Distant Supervision","year":2012,"lang":"en","type":"article","venue":"","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":24,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Blogosphere; Computer science; Classifier (UML); Vocabulary; Binary classification; Artificial intelligence; Class (philosophy); Information retrieval; Sample (material); Natural language processing; Data science; World Wide Web; Support vector machine; The Internet","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":[0.001392107,0.001003216,0.0007023741,0.002822393,0.0008039457,0.001036542,0.0008585103,0.001092456,0.002033274],"category_scores_gemma":[0.004580949,0.0002333489,0.0005752194,0.001771015,0.0004999064,0.001768907,0.001070891,0.0009529831,0.002646011],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004167375,"about_ca_system_score_gemma":0.0006471512,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003588024,"about_ca_topic_score_gemma":0.007169147,"domain_scores_codex":[0.9991001,0.0002245035,0.00007120199,0.000290744,0.0002101182,0.0001034152],"domain_scores_gemma":[0.9968811,0.001293401,0.0003568528,0.0004958877,0.0008362844,0.0001364936],"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":[0.001283424,0.0009871774,0.03822216,0.0004544263,0.0002338855,0.0003176863,0.0005960105,0.0240336,0.04082884,0.001838912,0.02847577,0.8627281],"study_design_scores_gemma":[0.0001159646,0.0004067787,0.03452908,0.00008289128,0.0001355259,0.0003527655,0.0003972759,0.9094405,0.03858063,0.007261618,0.008621554,0.00007536764],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.517301,0.002538458,0.4456151,0.0007696109,0.0004605144,0.0005640246,0.004519244,0.01222484,0.0160072],"genre_scores_gemma":[0.890008,0.0003679311,0.09284701,0.000113684,0.0003959351,0.0002116915,0.008201214,0.0002031662,0.007651242],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.003588024,"threshold_uncertainty_score":0.007362247,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06051903039669124,"score_gpt":0.3021238606092946,"score_spread":0.2416048302126033,"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."}}