{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0002857601,0.00021949,0.0002755779,0.000167962,0.0002513267,0.0002957958,0.0002377646,0.0001193999,0.000003525089],"category_scores_gemma":[0.00001826949,0.0001808242,0.0000425218,0.0003105009,0.00002900531,0.0005014943,0.00004496597,0.0002775183,9.188806e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003791372,"about_ca_system_score_gemma":0.00002015494,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000001385882,"about_ca_topic_score_gemma":1.505495e-7,"domain_scores_codex":[0.9989295,0.000009401183,0.0002781176,0.000355209,0.0001322732,0.000295492],"domain_scores_gemma":[0.9994366,0.000143577,0.00007750448,0.00021982,0.00007018482,0.00005232496],"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":[0.0000123716,0.00002910617,0.000001567112,0.0003935494,0.0000238514,2.415116e-7,0.0003107949,0.9729754,0.0002422195,0.01249686,0.000003065922,0.01351103],"study_design_scores_gemma":[0.0007354204,0.00008469007,0.000001933114,0.0005171623,0.00002425383,0.000009709674,0.00004093577,0.9943811,0.002322061,0.001511873,0.0001497289,0.0002211333],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.000937806,0.0001313871,0.9966506,0.0007429372,0.00007311183,0.0004299206,9.956798e-7,0.0007515462,0.0002817344],"genre_scores_gemma":[0.217564,0.0001226534,0.7817507,0.00001082945,0.000009953207,0.0001681813,0.00001015189,0.00001749964,0.0003460107],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.2166262,"threshold_uncertainty_score":0.7373795,"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."}}