{"id":"W2286678840","doi":"10.1177/1079063215616818","title":"Assessing the Validity of Automated Webcrawlers as Data Collection Tools to Investigate Online Child Sexual Exploitation","year":2015,"lang":"en","type":"article","venue":"Sexual Abuse","topic":"Sex work and related issues","field":"Social Sciences","cited_by":29,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University","funders":"","keywords":"The Internet; Data collection; Computer science; Distribution (mathematics); Selection (genetic algorithm); Data science; Artificial intelligence; World Wide Web; Statistics; Mathematics","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":[0.001589798,0.0001266808,0.0001763198,0.00009682066,0.00059309,0.0002696813,0.000547825,0.0001463326,0.00002802594],"category_scores_gemma":[0.002163665,0.00009554013,0.00001468819,0.001091864,0.0002681318,0.0009619589,0.0001029844,0.0002068826,0.00006799846],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001030063,"about_ca_system_score_gemma":0.0004063746,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003776613,"about_ca_topic_score_gemma":0.001736924,"domain_scores_codex":[0.9979807,0.0005433415,0.0002988667,0.000317751,0.0005906911,0.0002686337],"domain_scores_gemma":[0.9986604,0.0002602938,0.000162258,0.0004451892,0.0002446859,0.0002271424],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"qualitative","study_design_gemma":"qualitative","study_design_scores_codex":[0.00009348389,0.0006333949,0.01335908,0.00003623574,0.000206027,0.00001615174,0.8812283,0.003555342,0.003290882,0.0004209646,0.08112886,0.0160313],"study_design_scores_gemma":[0.001039659,0.000595419,0.01330744,0.0001467745,0.0001473135,0.000008359572,0.8845133,0.01237972,0.001147364,0.0007384916,0.08544512,0.0005310162],"study_design_candidate":"qualitative","study_design_consensus":"qualitative","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9928495,0.00007617493,0.00006368344,0.002971625,0.000847487,0.0003765321,0.00001511938,0.0003780198,0.002421806],"genre_scores_gemma":[0.995818,0.00003524887,0.001452728,0.0004397018,0.0004073909,0.000008401178,0.0002361422,0.00001929285,0.001583111],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01550028,"threshold_uncertainty_score":0.5709134,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2395439507976426,"score_gpt":0.4239995210445363,"score_spread":0.1844555702468937,"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."}}