{"id":"W4414693363","doi":"10.1109/tts.2025.3611984","title":"Using Large Language Models in Cluster Analysis in the Social Sciences","year":2025,"lang":"en","type":"article","venue":"IEEE Transactions on Technology and Society","topic":"Computational and Text Analysis Methods","field":"Social Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Institute on Governance","funders":"","keywords":"Cluster analysis; Metadata; Workflow; Context (archaeology); Variety (cybernetics); Exploratory data analysis; Thematic map; Topic model; Language model","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":[0.022795,0.001742003,0.001480987,0.006527554,0.002910902,0.006107543,0.003503267,0.001960278,0.002270716],"category_scores_gemma":[0.07492498,0.001019425,0.00354556,0.007553644,0.003515502,0.004580143,0.005417616,0.003797052,0.001364891],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.004241923,"about_ca_system_score_gemma":0.005754108,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02090693,"about_ca_topic_score_gemma":0.02901154,"domain_scores_codex":[0.9744922,0.01986177,0.0009152344,0.00263307,0.001725502,0.0003722157],"domain_scores_gemma":[0.9434264,0.04747707,0.001773811,0.004407293,0.002439179,0.0004762323],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0003592868,0.0003779976,0.01643511,0.0008432372,0.001387479,0.0004589553,0.006850435,0.4452344,0.002755708,0.2733958,0.0133038,0.2385978],"study_design_scores_gemma":[0.0000307243,0.00003095092,0.001179561,0.00007820116,0.00004538219,0.00004675468,0.0007126988,0.7444593,0.0008117354,0.2477913,0.004739593,0.00007378813],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.00516754,0.0001892966,0.9920541,0.0007274525,0.00005093339,0.0001567192,0.0003114543,0.0008155158,0.0005268932],"genre_scores_gemma":[0.1160212,0.0002737803,0.8790997,0.0003721949,0.0001432318,0.001311561,0.001349777,0.0004887506,0.0009399203],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.022795,"threshold_uncertainty_score":0.1205529,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04261437470856897,"score_gpt":0.4023507107515543,"score_spread":0.3597363360429853,"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."}}