{"id":"W7056428510","doi":"","title":"Enhancing performance of big data applying similarity over detected community","year":2023,"lang":"en","type":"dissertation","venue":"National Repository of Dissertations in Serbia","topic":"Magnetic confinement fusion research","field":"Physics and Astronomy","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Jaccard index; Similarity (geometry); Big data; Cosine similarity; Noise (video); Preprocessor; Social network (sociolinguistics); Sampling (signal processing); The Internet; Data pre-processing; Semantic similarity","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0007006973,0.0002016751,0.0003312239,0.0004801496,0.0002915331,0.00002895257,0.0007828042,0.000128542,0.001132119],"category_scores_gemma":[0.0002461965,0.0002337441,0.00007932364,0.000756233,0.00005340205,0.00007576623,0.0001280961,0.0006642208,0.000004862535],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003022812,"about_ca_system_score_gemma":0.0007270826,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00672247,"about_ca_topic_score_gemma":0.01015794,"domain_scores_codex":[0.9974435,0.0001708074,0.000905502,0.0002955576,0.0009956014,0.0001890298],"domain_scores_gemma":[0.9972134,0.0007138176,0.0005431658,0.0006368425,0.0008443829,0.0000484418],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"observational","study_design_scores_codex":[0.0005218686,0.001981187,0.3259946,0.003610491,0.0005340155,0.000003403261,0.008658193,0.0006004251,0.5360306,0.001755439,0.001177132,0.1191327],"study_design_scores_gemma":[0.0004840849,0.0000606385,0.9138513,0.0005605235,0.00005378173,1.794001e-7,0.008508713,0.003878243,0.07196049,0.0002478474,0.0001454382,0.0002487201],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9579796,0.00002586507,0.0000117315,0.000009375885,0.0003431221,0.0005093596,0.0001329345,0.0000197259,0.04096835],"genre_scores_gemma":[0.9851573,0.000005506784,0.0002705303,0.000001942855,0.00008292397,0.0001651827,0.006317334,0.0000268403,0.007972422],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5878568,"threshold_uncertainty_score":0.9998919,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05046243849428024,"score_gpt":0.3446469483686104,"score_spread":0.2941845098743302,"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."}}