{"id":"W2137811660","doi":"10.1140/epjds/s13688-014-0011-3","title":"Exploiting citation networks for large-scale author name disambiguation","year":2014,"lang":"en","type":"article","venue":"EPJ Data Science","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":53,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Social Sciences and Humanities Research Council of Canada","keywords":"Metadata; Ranking (information retrieval); Citation; Similarity (geometry); Cluster analysis; Precision and recall; Bibliographic coupling; Parameterized complexity; Heuristics","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["bibliometrics"],"consensus_categories":[],"category_scores_codex":[0.007875551,0.00151581,0.001648311,0.0156743,0.002226655,0.004066256,0.002725703,0.0026399,0.001089376],"category_scores_gemma":[0.05275797,0.0006727188,0.001279095,0.01324587,0.001183139,0.005235686,0.003091918,0.001735655,0.001595353],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001692871,"about_ca_system_score_gemma":0.002503476,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004798748,"about_ca_topic_score_gemma":0.006249134,"domain_scores_codex":[0.9936146,0.002078652,0.0005431711,0.001356441,0.002102727,0.0003043936],"domain_scores_gemma":[0.9688808,0.0195065,0.004609577,0.003269968,0.003267845,0.0004653209],"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":[0.0003155188,0.0002879369,0.04776144,0.0004870378,0.0005465207,0.0003848722,0.0005664773,0.4262884,0.007856273,0.03586206,0.008151426,0.471492],"study_design_scores_gemma":[0.00001548359,0.00002850881,0.002720439,0.00003833892,0.00003753697,0.0001452395,0.0000737966,0.9489338,0.004494795,0.03970053,0.00377303,0.00003852673],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.04042672,0.001421307,0.9515778,0.0004634721,0.00010868,0.0001049281,0.000964997,0.00299444,0.001937679],"genre_scores_gemma":[0.5006384,0.0007564305,0.492569,0.0001565203,0.000348369,0.0002271019,0.003282305,0.0003125059,0.001709287],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9843257,"threshold_uncertainty_score":0.04165035,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3376502821107572,"score_gpt":0.4681113964519575,"score_spread":0.1304611143412003,"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."}}