{"id":"W4406937501","doi":"10.1007/s10502-025-09475-z","title":"Conceptualizing aggregate-level description in web archives","year":2025,"lang":"en","type":"article","venue":"Archival Science","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Social Sciences and Humanities Research Council; Social Sciences and Humanities Research Council of Canada","keywords":"Cultural heritage; Web application; Aggregate (composite); World Wide Web; History; Data science; Geography; Computer science; Archaeology","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":[],"consensus_categories":[],"category_scores_codex":[0.01117757,0.0004705231,0.0006184735,0.01043712,0.004990531,0.01958934,0.002190756,0.002358295,0.003047179],"category_scores_gemma":[0.01958325,0.0006149028,0.0009357,0.01383776,0.01759296,0.03417632,0.009209013,0.002898134,0.0005257359],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.005915912,"about_ca_system_score_gemma":0.004158668,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01135402,"about_ca_topic_score_gemma":0.007326429,"domain_scores_codex":[0.9900981,0.005177719,0.001164653,0.0009229651,0.002107088,0.0005294851],"domain_scores_gemma":[0.9845459,0.007995497,0.001699844,0.003673818,0.001570043,0.0005149255],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.000009948252,0.000009042873,0.002164798,0.00006800646,0.0000123406,0.0001388573,0.02229303,0.000858038,0.0001458123,0.9644979,0.0004999031,0.009302219],"study_design_scores_gemma":[0.000009321625,0.0000243366,0.002261646,0.0003991664,0.00005195294,0.0005548591,0.03758712,0.008512066,0.0006610035,0.8494827,0.100415,0.00004077701],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1341923,0.004882377,0.7006361,0.01251937,0.0002823545,0.0004586569,0.0009829531,0.0007498305,0.1452961],"genre_scores_gemma":[0.8307077,0.00229926,0.1564591,0.0005207423,0.0001857818,0.0005402894,0.0009556909,0.000226748,0.008104793],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01958934,"threshold_uncertainty_score":0.05911332,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03931571689309058,"score_gpt":0.2820524073915678,"score_spread":0.2427366904984773,"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."}}