{"id":"W6920336244","doi":"10.60692/mf63k-29160","title":"Hash data files for \"Hashes are not suitable to verify fixity of the public archived web\"","year":2022,"lang":"en","type":"article","venue":"Greater South Information System","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Columbia College","funders":"","keywords":"Hash function; SHA-2; Merkle tree; Hash tree; Construct (python library); Set (abstract data type); Hash chain; Double hashing","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":[0.0008755169,0.0001070518,0.0002058878,0.0001936308,0.0004550632,0.0003515528,0.002652416,0.0000250137,0.00001626523],"category_scores_gemma":[0.0001134023,0.00007827627,0.00007773779,0.0004410074,0.00001957008,0.001203591,0.002103997,0.0000868206,0.00005512026],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004956731,"about_ca_system_score_gemma":0.0001042493,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001548019,"about_ca_topic_score_gemma":0.00000285459,"domain_scores_codex":[0.9985762,0.0001158308,0.0004452736,0.0002181165,0.0004207027,0.0002239163],"domain_scores_gemma":[0.9976845,0.00003369815,0.0003714038,0.001737545,0.0001078902,0.00006497359],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0002171949,0.00007355506,0.4937582,0.002695404,0.0008184375,0.000005811329,0.3062058,0.01177045,0.0001303346,0.01531522,0.154363,0.0146466],"study_design_scores_gemma":[0.00162293,0.0001421482,0.07552171,0.0002079486,0.00008454882,0.00003344501,0.03477857,0.7867187,0.001100564,0.000009796303,0.09916755,0.0006121607],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4641436,0.000004304549,0.519469,0.002557157,0.001029303,0.0008669475,0.01056065,0.0002864914,0.00108252],"genre_scores_gemma":[0.9960945,2.774262e-8,0.003080856,0.0002984625,0.00002786595,0.0001024737,0.0001703522,0.000003974416,0.000221503],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7749482,"threshold_uncertainty_score":0.4928894,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1013182629852221,"score_gpt":0.2350507976600053,"score_spread":0.1337325346747832,"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."}}