{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.04743181,0.0004774354,0.0005162551,0.01027458,0.001347189,0.003467527,0.001210929,0.0007716534,0.001928949],"category_scores_gemma":[0.2177048,0.0003512119,0.0006780405,0.006815125,0.001136328,0.004459372,0.002519733,0.0008452294,0.000894787],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001392333,"about_ca_system_score_gemma":0.002294033,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006172785,"about_ca_topic_score_gemma":0.01123607,"domain_scores_codex":[0.9614251,0.02645827,0.003182233,0.002141752,0.005898531,0.0008940402],"domain_scores_gemma":[0.602836,0.2906445,0.03332493,0.0260155,0.04420896,0.002970178],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0004924303,0.0006096801,0.8784127,0.0005604022,0.0002643238,0.0001106849,0.01398326,0.002130337,0.001468755,0.003178909,0.002444116,0.09634443],"study_design_scores_gemma":[0.0001170217,0.001258561,0.9032949,0.0006310757,0.0002055602,0.0003303179,0.0228447,0.04605669,0.006540598,0.006195086,0.01238913,0.0001363712],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.970647,0.0001998457,0.01701914,0.000235346,0.00005322379,0.0019825,0.001717344,0.0001860913,0.007959451],"genre_scores_gemma":[0.9640127,0.0001233746,0.03025644,0.0001019632,0.0000298115,0.002439308,0.001875421,0.00005315807,0.001107781],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.04743181,"threshold_uncertainty_score":0.2508464,"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."}}