{"id":"W2034194047","doi":"10.7202/1015788ar","title":"Mining the ‘Internet Graveyard’: Rethinking the Historians’ Toolkit","year":2013,"lang":"en","type":"article","venue":"Journal of the Canadian Historical Association","topic":"Digital and Traditional Archives Management","field":"Arts and Humanities","cited_by":20,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal","funders":"","keywords":"Context (archaeology); The Internet; Computer science; Cloud computing; Process (computing); World Wide Web; Digital library; Data science; Literacy; History; Archaeology; Political science; Law; Art","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":true,"about_ca":true,"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":[0.01228036,0.0004601284,0.0007526575,0.01154694,0.003599525,0.01496243,0.003397429,0.00159261,0.005814499],"category_scores_gemma":[0.05471964,0.0008641186,0.0007314717,0.01401576,0.01002975,0.02232785,0.009187802,0.003731478,0.001573714],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.004336901,"about_ca_system_score_gemma":0.009518312,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.06012792,"about_ca_topic_score_gemma":0.09441665,"domain_scores_codex":[0.9940782,0.002705671,0.0004812971,0.0007927956,0.001726423,0.0002156316],"domain_scores_gemma":[0.9645827,0.02018781,0.001248175,0.009299773,0.003759774,0.0009218785],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"qualitative","study_design_scores_codex":[0.00006787262,0.00005238291,0.01909027,0.0004471561,0.00007261684,0.0005425685,0.02836976,0.00715582,0.000817386,0.6084343,0.04249285,0.292457],"study_design_scores_gemma":[0.00002373405,0.00001536832,0.005777045,0.0006525478,0.00003486102,0.0003462754,0.01459546,0.03842016,0.002014252,0.48667,0.4513595,0.00009069301],"study_design_candidate":"qualitative","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.08552755,0.006280732,0.7497977,0.05775781,0.0008075549,0.0003299515,0.004749218,0.007095015,0.08765439],"genre_scores_gemma":[0.3320513,0.003732269,0.6464843,0.001755908,0.0003257337,0.0002004145,0.003212184,0.001874818,0.01036305],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.06012792,"threshold_uncertainty_score":0.1195559,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02613261404807319,"score_gpt":0.169156771675358,"score_spread":0.1430241576272848,"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."}}