{"id":"W2103528122","doi":"10.1145/1458082.1458194","title":"Efficient techniques for document sanitization","year":2008,"lang":"en","type":"article","venue":"","topic":"Advanced Steganography and Watermarking Techniques","field":"Computer Science","cited_by":89,"is_retracted":false,"has_abstract":true,"ca_institutions":"ASTER","funders":"","keywords":"Computer science; Confidentiality; Information retrieval; World Wide Web; Computer security","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":[0.001458744,0.001370562,0.001529284,0.002502482,0.001214871,0.001664596,0.001902698,0.001491873,0.003628935],"category_scores_gemma":[0.005558532,0.0007718373,0.001104388,0.001709298,0.001481987,0.00470741,0.002653775,0.002137298,0.002133262],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005876365,"about_ca_system_score_gemma":0.001069504,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0007543697,"about_ca_topic_score_gemma":0.001203907,"domain_scores_codex":[0.9974195,0.0003934166,0.0001706859,0.0003946352,0.001387334,0.0002343762],"domain_scores_gemma":[0.9951242,0.001817515,0.0005994107,0.001803246,0.0005872032,0.00006840358],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0003310778,0.0002461549,0.0006319559,0.0004839273,0.0001078009,0.0001716942,0.0004200686,0.04104328,0.1177824,0.03817883,0.005590709,0.7950121],"study_design_scores_gemma":[0.0002367541,0.0004256786,0.001122112,0.0001290574,0.0001520129,0.00218081,0.0004333749,0.6519135,0.2411794,0.06059406,0.04151455,0.0001186955],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.009152744,0.000889184,0.9862146,0.0001304094,0.00004279834,0.0001077799,0.00005176644,0.001871934,0.001538814],"genre_scores_gemma":[0.08987419,0.001227039,0.9044237,0.0000698891,0.00007124191,0.0001326869,0.0002474648,0.0002204905,0.003733228],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.003628935,"threshold_uncertainty_score":0.01214004,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01476314489143149,"score_gpt":0.2555428132089694,"score_spread":0.2407796683175379,"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."}}