{"id":"W4395666468","doi":"10.18280/mmep.110422","title":"Rule-Based Information Extraction from Multi-format Resumes for Automated Classification","year":2024,"lang":"en","type":"article","venue":"Mathematical Modelling and Engineering Problems","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Computer science; Information retrieval; Information extraction; Data mining","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"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":[0.0008821936,0.001236686,0.0008559451,0.006349136,0.0006477766,0.001765421,0.001306475,0.0008760468,0.005170716],"category_scores_gemma":[0.00440424,0.0002811537,0.001363887,0.00425118,0.0003532741,0.001850618,0.0007734071,0.00120887,0.006905326],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005194321,"about_ca_system_score_gemma":0.001335887,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004098319,"about_ca_topic_score_gemma":0.005569022,"domain_scores_codex":[0.9989184,0.0001176552,0.0002064759,0.000319874,0.000358117,0.0000795745],"domain_scores_gemma":[0.9976719,0.0008877437,0.0002618927,0.0003775538,0.0007297932,0.00007115902],"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":[0.0002847242,0.0004142875,0.00579442,0.0008731838,0.00009549262,0.001164805,0.0004044532,0.005216834,0.03652542,0.003152447,0.03271833,0.9133556],"study_design_scores_gemma":[0.0001602123,0.0006335244,0.04067468,0.001080296,0.0005734446,0.002856375,0.002743689,0.5208401,0.1855114,0.0270963,0.2174928,0.000337177],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.08112526,0.003739411,0.8233704,0.001203153,0.0007843863,0.001589947,0.02993803,0.03715777,0.02109164],"genre_scores_gemma":[0.1900113,0.00176459,0.7411034,0.000285373,0.0002608891,0.0007280371,0.05414264,0.000493559,0.01121023],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.006349136,"threshold_uncertainty_score":0.01729774,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03277580496070964,"score_gpt":0.2708636690440989,"score_spread":0.2380878640833892,"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."}}