{"id":"W3043085860","doi":"10.17613/34jx-fw32","title":"Seeing the Heiltsuk orthography from font encoding through to Unicode: A case study using convertextract","year":2018,"lang":"en","type":"article","venue":"Humanities Commons CORE (Modern Language Association / Columbia University)","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"Social Sciences and Humanities Research Council of Canada","keywords":"Unicode; Computer science; World Wide Web; Transliteration; Encoding (memory); Font; Orthography; Writing system; Rendering (computer graphics); Natural language processing; Artificial intelligence; Linguistics","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":[0.001609647,0.0006507858,0.0003989863,0.001085606,0.008029916,0.004472155,0.001428807,0.002633448,0.005889633],"category_scores_gemma":[0.007297273,0.0004826473,0.0003936045,0.001637122,0.004653058,0.004187164,0.002900437,0.003724404,0.001278709],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003248316,"about_ca_system_score_gemma":0.00232851,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02710075,"about_ca_topic_score_gemma":0.08784703,"domain_scores_codex":[0.9979759,0.0009502996,0.0001472373,0.0002721911,0.0004580315,0.0001962862],"domain_scores_gemma":[0.9975102,0.001588974,0.0001931994,0.0003081512,0.0002279809,0.0001714668],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"qualitative","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0001886636,0.0006341396,0.009905721,0.0004698957,0.00002146993,0.09153312,0.7223507,0.001002299,0.009990013,0.0335531,0.01027222,0.1200787],"study_design_scores_gemma":[0.00003646865,0.0003687472,0.0149244,0.0005136956,0.00006432788,0.09022726,0.5426828,0.004153156,0.02920773,0.009638157,0.307979,0.0002041719],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9290594,0.0005888471,0.009684894,0.002992673,0.0001574298,0.0001593845,0.0001883727,0.0002570936,0.05691192],"genre_scores_gemma":[0.9213325,0.00108545,0.02320834,0.0009124801,0.00003922741,0.00009736871,0.0001746535,0.0005720892,0.05257795],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.02710075,"threshold_uncertainty_score":0.053886,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0426305788822605,"score_gpt":0.2697616665529911,"score_spread":0.2271310876707305,"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."}}