{"id":"W4292794274","doi":"10.23919/annsim55834.2022.9859486","title":"Incremental Text Clustering Algorithm For Cloud-Based Data Management In Scientific Research Papers","year":2022,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Toronto Metropolitan University","funders":"Ryerson University","keywords":"Cluster analysis; Computer science; Word2vec; Data mining; Correlation clustering; CURE data clustering algorithm; Document clustering; Canopy clustering algorithm; Conceptual clustering; Artificial intelligence; Information retrieval; Embedding","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001660176,0.001084232,0.001281048,0.009806645,0.001908326,0.003007997,0.002418098,0.0009769337,0.003538396],"category_scores_gemma":[0.005910044,0.0004623677,0.001561966,0.01082229,0.000449056,0.002785968,0.001903193,0.0008993038,0.003965068],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001666162,"about_ca_system_score_gemma":0.004139427,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006904656,"about_ca_topic_score_gemma":0.009203685,"domain_scores_codex":[0.997792,0.0001858322,0.0002650076,0.0007300287,0.0008164771,0.0002106244],"domain_scores_gemma":[0.9961603,0.0005679569,0.0004961581,0.0006394163,0.001819605,0.0003165751],"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":[0.0004764994,0.0003867379,0.007926409,0.0004777994,0.0001929457,0.0003232058,0.0005532798,0.02116537,0.0192886,0.007510764,0.02989855,0.9117998],"study_design_scores_gemma":[0.0002737456,0.0003604739,0.01138738,0.0001306381,0.0002783183,0.001068126,0.001329721,0.8391948,0.04426133,0.02916033,0.07239776,0.0001574198],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05587086,0.002767268,0.9136922,0.0009255787,0.000793737,0.001172762,0.005464348,0.01377307,0.005540221],"genre_scores_gemma":[0.1173853,0.0008210258,0.867898,0.0001485805,0.0002923776,0.000599984,0.007746194,0.0003938611,0.004714579],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.009806645,"threshold_uncertainty_score":0.01372892,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1470932673754568,"score_gpt":0.3605991523184348,"score_spread":0.213505884942978,"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."}}