{"id":"W3120354622","doi":"10.1007/s42803-020-00029-6","title":"From archive to analysis: accessing web archives at scale through a cloud-based interface","year":2021,"lang":"en","type":"article","venue":"International Journal of Digital Humanities","topic":"Advanced Data Storage Technologies","field":"Computer Science","cited_by":16,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo; York University","funders":"University of Waterloo; York University; Compute Canada; Andrew W. Mellon Foundation","keywords":"Cloud computing; World Wide Web; Computer science; Interface (matter); Scope (computer science); Mashup; Scale (ratio); Web development; Web service; Web page","routes":{"ca_aff":true,"ca_fund":true,"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":[0.003458331,0.0004446219,0.0006422605,0.001771893,0.002097561,0.007266908,0.001592425,0.0009746221,0.008471142],"category_scores_gemma":[0.01201423,0.0004269315,0.0004423159,0.002279681,0.001799891,0.007390967,0.007271898,0.001536061,0.002309562],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00112046,"about_ca_system_score_gemma":0.002137359,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008551265,"about_ca_topic_score_gemma":0.01087154,"domain_scores_codex":[0.997621,0.000867496,0.0001939773,0.000359389,0.0007733548,0.0001847583],"domain_scores_gemma":[0.9894313,0.004793478,0.0004908395,0.002758462,0.0015369,0.000989142],"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":[0.002779109,0.000826242,0.04538291,0.001130781,0.0002339024,0.004497912,0.06447963,0.007844551,0.05880792,0.1346654,0.1211515,0.5582001],"study_design_scores_gemma":[0.000441241,0.0003813888,0.03013942,0.001105123,0.0002439589,0.00256462,0.02969324,0.1201064,0.03890477,0.1361962,0.6397519,0.0004718028],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.2300162,0.001273655,0.6391175,0.008832437,0.0004887432,0.001092945,0.002526542,0.0534369,0.06321509],"genre_scores_gemma":[0.7093561,0.0007440686,0.2665922,0.002087931,0.0002770826,0.0004241581,0.00132817,0.005604245,0.01358595],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.008551265,"threshold_uncertainty_score":0.02833873,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03119082774501818,"score_gpt":0.3008049511167233,"score_spread":0.2696141233717051,"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."}}