{"id":"W4414652086","doi":"10.31234/osf.io/v28nf_v1","title":"Extracting Prototypes from Lexical Feature Norms for Settlement Concepts","year":2025,"lang":"en","type":"preprint","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Social Sciences and Humanities Research Council of Canada","keywords":"Context (archaeology); Feature (linguistics); Human settlement; Settlement (finance); Task (project management); Categorization; Abstraction; Semantics (computer science)","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":true,"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":[0.002881995,0.0006612029,0.0005234917,0.003751965,0.0006863212,0.00262642,0.001122628,0.0009994388,0.002694186],"category_scores_gemma":[0.02332332,0.0005037949,0.0007821944,0.001902127,0.001195759,0.005712057,0.001814432,0.0009964871,0.0007555532],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001134426,"about_ca_system_score_gemma":0.001011937,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002411114,"about_ca_topic_score_gemma":0.002853981,"domain_scores_codex":[0.9974824,0.0009124227,0.0002275731,0.0006645795,0.0005753716,0.0001376318],"domain_scores_gemma":[0.9885693,0.006910334,0.001199474,0.001156179,0.001850724,0.0003139582],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001308773,0.000423273,0.08277544,0.002077749,0.0002334778,0.002479833,0.0748743,0.006595084,0.07131656,0.07549074,0.01130091,0.6711239],"study_design_scores_gemma":[0.0004559271,0.001167375,0.1217226,0.0009791601,0.0004072745,0.004345954,0.1005011,0.3238684,0.04159562,0.3288571,0.07553396,0.0005655129],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7551741,0.0003846591,0.2318326,0.0004184508,0.00009589397,0.0004545851,0.001618068,0.001429376,0.008592374],"genre_scores_gemma":[0.8894923,0.0001225717,0.1070391,0.00003478341,0.00002504237,0.0003414786,0.002059142,0.0001306141,0.0007550098],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.003751965,"threshold_uncertainty_score":0.01524162,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02277203231235816,"score_gpt":0.3546443799982356,"score_spread":0.3318723476858774,"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."}}