{"id":"W4415707048","doi":"10.1109/access.2025.3627187","title":"Boosting Arabic Fake Reviews Detection by Integrating Textual and Metadata Features: A Transformer-Based Model","year":2025,"lang":"en","type":"article","venue":"IEEE Access","topic":"Spam and Phishing Detection","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Toronto Metropolitan University","funders":"Shanghai University of Engineering and Science; Toronto Metropolitan University","keywords":"Metadata; Arabic; Boosting (machine learning); Language model; Variety (cybernetics); Modern Standard Arabic; User-generated content","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0005205474,0.0001710166,0.000218984,0.0001513906,0.0002990527,0.0009562068,0.0007061382,0.00009035933,0.000001140864],"category_scores_gemma":[0.000154851,0.0001434664,0.0000644086,0.0006089272,0.00002886229,0.002092995,0.00004384255,0.0002875275,0.000001592549],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004485109,"about_ca_system_score_gemma":0.00005158556,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002942575,"about_ca_topic_score_gemma":0.0007674478,"domain_scores_codex":[0.9988238,0.00009022453,0.000280546,0.0004410434,0.0001543703,0.0002099921],"domain_scores_gemma":[0.9993042,0.0001312809,0.0001045171,0.0003563763,0.00005071444,0.00005293865],"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":[0.00002395604,0.00003865098,0.0001271234,0.0001671041,0.00002391783,0.000001453814,0.0003385471,0.001224255,0.07936597,0.0007245947,0.002970543,0.9149939],"study_design_scores_gemma":[0.000529198,0.00007100937,0.0002267675,0.0002901882,0.00005193341,0.000008721177,0.0000292085,0.8370852,0.1532371,0.002076706,0.006067206,0.0003268078],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01558976,0.001537339,0.9808679,0.0003566266,0.0003755711,0.0002841779,0.000007522323,0.0001510538,0.0008300474],"genre_scores_gemma":[0.9928925,0.0001026634,0.005670579,0.0009783081,0.00004518133,0.0000593488,0.000004864713,0.000007988604,0.0002385915],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9773027,"threshold_uncertainty_score":0.9220724,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03002095029311277,"score_gpt":0.3085211267269561,"score_spread":0.2785001764338434,"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."}}