{"id":"W4399906408","doi":"10.18280/ria.380307","title":"A New Algorithm for Arabic Document Clustering Utilizing Maximal Wordsets","year":2024,"lang":"fr","type":"article","venue":"Revue d intelligence artificielle","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Arabic; Cluster analysis; Document clustering; Computer science; Artificial intelligence; Information retrieval; Algorithm; Natural language processing; Pattern recognition (psychology); Linguistics","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":[0.001446567,0.001895461,0.001517737,0.006582889,0.002026071,0.002071908,0.002209765,0.001183394,0.004473195],"category_scores_gemma":[0.004570691,0.000641189,0.001653546,0.005915607,0.0007499283,0.002500017,0.001596751,0.001416348,0.004573333],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001190407,"about_ca_system_score_gemma":0.002225915,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005594385,"about_ca_topic_score_gemma":0.006626624,"domain_scores_codex":[0.997847,0.0002535425,0.0003231664,0.0006135763,0.0008385561,0.0001242665],"domain_scores_gemma":[0.9985701,0.000387139,0.0001301435,0.0002192347,0.0006338045,0.00005967557],"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":[0.0002068782,0.00007934986,0.0007510377,0.0002100429,0.00010998,0.00009803289,0.0002434793,0.01870275,0.01056994,0.007228182,0.008175842,0.9536244],"study_design_scores_gemma":[0.0001569377,0.000300265,0.002161504,0.0001274279,0.0001422077,0.00116219,0.0005840982,0.8084873,0.0437475,0.05081768,0.09215004,0.0001628661],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.00424615,0.0004223326,0.9906501,0.000119601,0.0001408672,0.0002687373,0.0004788161,0.002226579,0.001446787],"genre_scores_gemma":[0.01794665,0.0002163573,0.977443,0.00006594235,0.00006946884,0.000292885,0.001259195,0.0001968373,0.002509612],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.006582889,"threshold_uncertainty_score":0.0149644,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07427204738448545,"score_gpt":0.3190446043352724,"score_spread":0.2447725569507869,"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."}}