{"id":"W2405071382","doi":"10.5281/zenodo.43767","title":"Novel Topic N-Gram Count Lm Incorporating Document-Based Topic Distributions And N-Gram Counts","year":2014,"lang":"en","type":"article","venue":"Zenodo (CERN European Organization for Nuclear Research)","topic":"Topic Modeling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Institut National de la Recherche Scientifique","funders":"","keywords":"Gram; n-gram; Statistics; Computer science; Mathematics; Natural language processing; Biology; Bacteria","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003098312,0.002046024,0.001973091,0.001419972,0.001387882,0.00220018,0.002587455,0.003362433,0.007723649],"category_scores_gemma":[0.009796181,0.0008839273,0.001361601,0.002007572,0.0005201421,0.003434614,0.002063317,0.003368167,0.009353274],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009291136,"about_ca_system_score_gemma":0.002546558,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005298848,"about_ca_topic_score_gemma":0.01226507,"domain_scores_codex":[0.9976546,0.001044243,0.0001657698,0.0005387329,0.0004407436,0.0001559153],"domain_scores_gemma":[0.9952857,0.003037636,0.0001406869,0.0004505653,0.000898745,0.0001867004],"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":[0.001231057,0.0003365419,0.001272486,0.0005989493,0.0004365181,0.000355143,0.0003266601,0.04847033,0.03367909,0.01258873,0.05116425,0.8495403],"study_design_scores_gemma":[0.00007172823,0.00009130774,0.0005330921,0.00003936122,0.00008942094,0.0001304267,0.00005399079,0.9748317,0.00761674,0.007365225,0.009131998,0.00004505287],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.009761942,0.001509163,0.9756797,0.0009580541,0.0008378896,0.0001368034,0.00127385,0.007575063,0.002267571],"genre_scores_gemma":[0.140901,0.001281827,0.8210356,0.001110644,0.001662194,0.001002416,0.009616937,0.00256102,0.02082851],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007723649,"threshold_uncertainty_score":0.0258382,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02703303707625021,"score_gpt":0.2396609902227834,"score_spread":0.2126279531465332,"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."}}