{"id":"W4324277966","doi":"10.1177/20539517231163172","title":"All WARC and no playback: The materialities of data-centered web archives research","year":2023,"lang":"en","type":"article","venue":"Big Data & Society","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":10,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Social Sciences and Humanities Research Council of Canada","keywords":"Computer science; Standardization; Metadata; World Wide Web; Interoperability; Data science","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["sts"],"consensus_categories":[],"category_scores_codex":[0.05241829,0.0005746887,0.0006344186,0.008180865,0.01138663,0.03386647,0.002102296,0.002413765,0.005812851],"category_scores_gemma":[0.09719202,0.001003984,0.0006685077,0.008507095,0.06463737,0.04559393,0.01557106,0.004386825,0.0008010112],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.006974788,"about_ca_system_score_gemma":0.007167382,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002334212,"about_ca_topic_score_gemma":0.002801642,"domain_scores_codex":[0.9553988,0.03443685,0.001434767,0.002764444,0.005016304,0.0009487402],"domain_scores_gemma":[0.8821685,0.08321582,0.004005249,0.02517539,0.00377727,0.001657891],"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":[0.0000771445,0.00003400837,0.002395082,0.000245829,0.00001535208,0.0002443152,0.1254015,0.0002455921,0.0003756211,0.8171774,0.002813624,0.05097456],"study_design_scores_gemma":[0.00003811871,0.0001003217,0.00256484,0.001858504,0.00004095536,0.0008469473,0.1250915,0.001003284,0.002007845,0.5847985,0.281563,0.00008629145],"study_design_candidate":"qualitative","study_design_consensus":null,"genre_codex":"other","genre_gemma":"empirical","genre_scores_codex":[0.2270503,0.01457466,0.3122047,0.09902659,0.002484314,0.0006870534,0.0008031867,0.001099285,0.3420699],"genre_scores_gemma":[0.9104668,0.003691031,0.06929074,0.004080527,0.0008195897,0.0007913467,0.0003103751,0.0008371609,0.009712418],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9886134,"threshold_uncertainty_score":0.2772177,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4562295496170037,"score_gpt":0.3975399753386789,"score_spread":0.05868957427832477,"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."}}