{"id":"W2773853552","doi":"10.2200/s00809ed2v01y201710hlt038","title":"Natural Language Processing for Social Media, Second Edition","year":2017,"lang":"en","type":"article","venue":"Synthesis lectures on human language technologies","topic":"Sentiment Analysis and Opinion Mining","field":"Computer Science","cited_by":19,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Ottawa","funders":"University of Ottawa; University of Toronto; University of Southern California","keywords":"Social media; Linguistics; Computer science; Natural language processing; World Wide Web; Philosophy","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":"codex-gemma-dda1882f352a","candidate_categories":["sts"],"consensus_categories":[],"category_scores_codex":[0.0002469122,0.0002137812,0.0002821791,0.0002889438,0.001470921,0.0007466141,0.001706164,0.0001748121,0.00007329124],"category_scores_gemma":[0.001032146,0.0001714898,0.0001697129,0.00009987739,0.0001837836,0.0003933221,0.0002508397,0.0002215155,0.00001168134],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004499787,"about_ca_system_score_gemma":0.00001771735,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000007017219,"about_ca_topic_score_gemma":0.0001383897,"domain_scores_codex":[0.9986998,0.0000270191,0.0001982641,0.0004644478,0.0002669642,0.0003435095],"domain_scores_gemma":[0.998622,0.0002631562,0.0003075091,0.0007382414,0.00004969373,0.00001937232],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0000132694,0.00005619516,0.00005370958,0.0000665343,0.00007501985,0.00002689301,0.007588515,0.000002001625,0.03668664,0.009032653,0.005720583,0.940678],"study_design_scores_gemma":[0.0008890888,0.0001323479,0.00722944,0.0002686632,0.0001035627,0.0000103266,0.01494332,0.004766916,0.9577296,0.01012837,0.002575361,0.001222986],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9460585,0.007614503,0.02480303,0.007791613,0.001408967,0.0008340177,0.00006960653,0.005638001,0.005781732],"genre_scores_gemma":[0.994769,0.000003206132,0.004471914,0.00009308009,0.0003346945,0.0001278277,0.00001322974,0.00001725589,0.000169772],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.939455,"threshold_uncertainty_score":0.9998291,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03585718823322134,"score_gpt":0.3292590273087216,"score_spread":0.2934018390755003,"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."}}