{"id":"W4288060620","doi":"10.18357/kula.221","title":"Re-purposing Excavation Database Content as Paradata","year":2022,"lang":"en","type":"article","venue":"KULA knowledge creation dissemination and preservation studies","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":35,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Riksbankens Jubileumsfond; European Commission","keywords":"Metadata; Computer science; Documentation; Field (mathematics); Identification (biology); Structuring; Process (computing); Data science; Excavation; Reading (process); Plan (archaeology); Archaeology; World Wide Web","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.02410499,0.001003111,0.0009090451,0.01664614,0.002599229,0.01311983,0.002576896,0.001167123,0.004294188],"category_scores_gemma":[0.08772025,0.001167313,0.001107124,0.01789346,0.003469078,0.01283696,0.008262751,0.002266109,0.003532749],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003208768,"about_ca_system_score_gemma":0.006090172,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006478499,"about_ca_topic_score_gemma":0.01029973,"domain_scores_codex":[0.9786617,0.006428795,0.004231226,0.003298228,0.006623104,0.0007570577],"domain_scores_gemma":[0.8662123,0.05790285,0.006739152,0.04485409,0.02316818,0.001123394],"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":[0.000519445,0.0002244238,0.04022616,0.004683125,0.0001759139,0.00165496,0.1504841,0.003137483,0.02350389,0.1291157,0.03591946,0.6103554],"study_design_scores_gemma":[0.00004292636,0.0001098648,0.02812449,0.002747046,0.0001494333,0.001185337,0.06018271,0.008592186,0.02106653,0.06723518,0.8103509,0.0002134391],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1743659,0.002022158,0.6719247,0.006397971,0.001121836,0.005219096,0.0576246,0.007761756,0.07356205],"genre_scores_gemma":[0.317512,0.001125814,0.6179843,0.0009581641,0.0002464501,0.003674174,0.03845707,0.004395445,0.0156466],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9868802,"threshold_uncertainty_score":0.1274809,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09820098497722425,"score_gpt":0.3918604070508239,"score_spread":0.2936594220735996,"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."}}