{"id":"W2578945560","doi":"","title":"Real Time Filtering of Tweets Using Wikipedia Concepts and Google Tri-gram Semantic Relatedness","year":2015,"lang":"en","type":"article","venue":"Text REtrieval Conference","topic":"Topic Modeling","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University","funders":"","keywords":"Computer science; Information retrieval; n-gram; Semantic similarity; Word (group theory); Social media; World Wide Web; Natural language processing; Language model","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":[0.001019977,0.001464406,0.001045533,0.01125887,0.001144069,0.001429148,0.0006428831,0.0008962401,0.001694818],"category_scores_gemma":[0.005173184,0.0002693048,0.001027007,0.005165648,0.0003010747,0.002671326,0.0009942116,0.0007721198,0.002620385],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007016975,"about_ca_system_score_gemma":0.0009506127,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01083807,"about_ca_topic_score_gemma":0.01624328,"domain_scores_codex":[0.9986695,0.0002659899,0.0001376662,0.0003681298,0.0004247991,0.0001339713],"domain_scores_gemma":[0.9982017,0.0005624624,0.0002036956,0.0001599977,0.0007584417,0.0001135743],"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":[0.002362785,0.00142356,0.02747598,0.00170989,0.0005965793,0.001430453,0.002094018,0.01695572,0.10253,0.005449547,0.0458358,0.7921356],"study_design_scores_gemma":[0.0001348757,0.0009880048,0.04634593,0.0001829496,0.0003468475,0.001177295,0.00245571,0.8449103,0.05352491,0.01233835,0.03733975,0.0002550446],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5646859,0.003639693,0.3740354,0.001137115,0.0009863785,0.001555434,0.01720816,0.02322814,0.01352379],"genre_scores_gemma":[0.6769998,0.0007001322,0.2939443,0.000160422,0.000344566,0.0005274453,0.0233479,0.0004239961,0.003551478],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01125887,"threshold_uncertainty_score":0.02154994,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07837344284233229,"score_gpt":0.3094833050650553,"score_spread":0.231109862222723,"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."}}