{"id":"W4414360013","doi":"10.18280/mmep.120835","title":"Enhanced Text Extraction: Combining Bacteria Foraging Optimization Algorithm-Optimized Scale-Invariant Feature Transform with Machine Learning for Robust Performance","year":2025,"lang":"en","type":"article","venue":"Mathematical Modelling and Engineering Problems","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Pattern recognition (psychology); Feature (linguistics); Support vector machine; Feature selection; Foraging; Training set","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":[0.0005596736,0.0009208631,0.0009767998,0.00115987,0.0002935267,0.0007612152,0.001133352,0.0008986108,0.002098483],"category_scores_gemma":[0.0015962,0.0002495454,0.0008063191,0.001061429,0.0003097503,0.001428621,0.0005717317,0.0006574704,0.002295333],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004674436,"about_ca_system_score_gemma":0.0005933215,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002525163,"about_ca_topic_score_gemma":0.003084454,"domain_scores_codex":[0.9996496,0.00003626434,0.00002331686,0.00008590111,0.0001639551,0.00004088664],"domain_scores_gemma":[0.9996785,0.00008757425,0.00004488974,0.00005210321,0.0001211243,0.00001584055],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0002562432,0.0001519242,0.00109879,0.0001463337,0.0001063349,0.0001544521,0.00005684794,0.09214259,0.1004928,0.003131034,0.007008587,0.795254],"study_design_scores_gemma":[0.00001821689,0.00005861302,0.0004679437,0.00000689984,0.00001912596,0.00008904621,0.00001270117,0.9735019,0.02207399,0.001480932,0.002254771,0.00001600355],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01917002,0.000489237,0.9754974,0.0001317986,0.00007529839,0.00005894156,0.0001173843,0.003128878,0.001331075],"genre_scores_gemma":[0.2941267,0.000390029,0.6974141,0.0002509987,0.0001215026,0.0001573068,0.0008982326,0.00055145,0.006089709],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.002525163,"threshold_uncertainty_score":0.007020056,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01353800708907852,"score_gpt":0.2110719321133081,"score_spread":0.1975339250242296,"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."}}