{"id":"W4410794822","doi":"10.1177/08944393251344865","title":"Prompting the Machine: Introducing an LLM Data Extraction Method for Social Scientists","year":2025,"lang":"en","type":"article","venue":"Social Science Computer Review","topic":"Topic Modeling","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal; Université Laval","funders":"","keywords":"Computer science; Extraction (chemistry); Data extraction; Data science; Political science; MEDLINE; Chromatography; Chemistry","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["sts","open_science"],"consensus_categories":[],"category_scores_codex":[0.01016946,0.0001750547,0.000321806,0.000119198,0.002886809,0.0009191422,0.005796478,0.00004663135,0.000004265964],"category_scores_gemma":[0.0002836125,0.0001319629,0.0001004761,0.002041955,0.0002479736,0.001991174,0.002130746,0.0002439408,0.000004963241],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001528383,"about_ca_system_score_gemma":0.0004980013,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004864556,"about_ca_topic_score_gemma":0.00001771043,"domain_scores_codex":[0.9967906,0.0003257209,0.0004658294,0.001271928,0.0006106774,0.0005352036],"domain_scores_gemma":[0.9980224,0.0001720626,0.0002526924,0.001202989,0.0002826302,0.00006719265],"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":[6.770169e-7,0.00003115284,0.00001176265,0.0002509282,0.000005731426,4.607047e-7,0.0005459064,0.00001914117,0.0001746651,0.07366003,0.002373819,0.9229257],"study_design_scores_gemma":[0.0001614221,0.00002537616,0.0005862593,0.0004070819,0.0000392919,0.00001043369,0.00002175953,0.9245062,0.00004855783,0.005476851,0.06848115,0.0002355977],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0001101679,0.002671344,0.9794277,0.01443031,0.001889933,0.0008877537,0.000005547749,0.000142089,0.0004351331],"genre_scores_gemma":[0.01212357,0.0002284569,0.9793918,0.006081105,0.001942741,0.00008966786,0.00001885257,0.00001276748,0.0001109956],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9244871,"threshold_uncertainty_score":0.9995826,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1056737097266691,"score_gpt":0.46011694349039,"score_spread":0.3544432337637208,"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."}}