{"id":"W4409813933","doi":"10.1016/j.procs.2025.03.116","title":"Harnessing Large Language Models for Precision Topic Extraction and Technology Patent Nomination: A GPT-centric Methodology","year":2025,"lang":"en","type":"article","venue":"Procedia Computer Science","topic":"Advanced Text Analysis Techniques","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Cégep de Rimouski; Université du Québec à Rimouski","funders":"","keywords":"Computer science; Nomination; Patent analysis; Extraction (chemistry); Data science; Data mining","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00347375,0.001323236,0.0009218327,0.003983257,0.0008704234,0.002797853,0.001241963,0.001309545,0.00191317],"category_scores_gemma":[0.01746302,0.0005086447,0.001909668,0.00344139,0.0007572367,0.0041606,0.001805459,0.002215553,0.002122995],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001051468,"about_ca_system_score_gemma":0.002148942,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003541529,"about_ca_topic_score_gemma":0.005370729,"domain_scores_codex":[0.9976399,0.001166286,0.0001473533,0.0005719155,0.0003675486,0.0001070025],"domain_scores_gemma":[0.9876809,0.009922191,0.0008007218,0.0006981569,0.0007474538,0.0001506033],"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":[0.0004068278,0.0004300242,0.01207133,0.00101586,0.0005602521,0.0007024991,0.002449128,0.15131,0.02005783,0.06597355,0.01494133,0.7300814],"study_design_scores_gemma":[0.00003526331,0.00008111005,0.001547195,0.00006244336,0.0001105443,0.0001959786,0.0002345417,0.9311917,0.004232899,0.05318401,0.009075004,0.00004937227],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0195126,0.0008461376,0.9744221,0.0009259254,0.0000819682,0.0001927555,0.0009293712,0.00136622,0.00172291],"genre_scores_gemma":[0.4630515,0.001531889,0.5227875,0.0005407681,0.0006495398,0.001018782,0.005334745,0.0004368802,0.004648441],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.003983257,"threshold_uncertainty_score":0.01837111,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04369533804685807,"score_gpt":0.3513519287723578,"score_spread":0.3076565907254998,"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."}}