{"id":"W4205471046","doi":"10.1093/llc/fqab108","title":"An approach to complex texts in multiple documents","year":2022,"lang":"en","type":"article","venue":"Digital Scholarship in the Humanities","topic":"Digital Humanities and Scholarship","field":"Arts and Humanities","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Saskatchewan","funders":"Arts and Humanities Research Council","keywords":"Computer science; Inscribed figure; Hierarchy; Architecture; World Wide Web; Transcription (linguistics); JSON; Information retrieval; Linguistics; History","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","sts","scholarly_communication","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.001195077,0.0003708082,0.0003456755,0.0005414402,0.001694881,0.008630112,0.001891438,0.00003855619,0.001414732],"category_scores_gemma":[0.0001031843,0.0003202196,0.0001292128,0.0002084902,0.0003947458,0.004566151,0.0004129036,0.0009117134,0.0001094463],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002689804,"about_ca_system_score_gemma":0.00004832761,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0004065753,"about_ca_topic_score_gemma":0.002799522,"domain_scores_codex":[0.9968323,0.000362974,0.0005906206,0.0005315045,0.0009315475,0.0007510357],"domain_scores_gemma":[0.9988902,0.0002204717,0.0001056327,0.0006241555,0.0000715641,0.00008799353],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0001689508,0.00192495,0.01599301,0.0000418063,0.00002213973,0.00006546897,0.186786,0.0004192124,0.000009352195,0.7898937,0.001551382,0.003124077],"study_design_scores_gemma":[0.001185157,0.0005963641,0.03529425,0.00003888214,0.000009709928,0.00003218504,0.1883916,0.00009482978,0.00000334839,0.05294142,0.7205239,0.0008883912],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5801086,0.00007527619,0.000001232459,0.00007298381,0.0001943949,0.0005256105,0.0003736527,0.00008258792,0.4185656],"genre_scores_gemma":[0.9853958,0.000001063235,0.0000157891,0.003698045,0.000263035,0.0005281545,0.0004558696,0.0000589901,0.009583257],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7369522,"threshold_uncertainty_score":0.999925,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1365965162691597,"score_gpt":0.2733653513359369,"score_spread":0.1367688350667772,"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."}}