{"id":"W2963665652","doi":"10.3390/bdcc3030044","title":"Archetype-Based Modeling and Search of Social Media","year":2019,"lang":"en","type":"article","venue":"Big Data and Cognitive Computing","topic":"Advanced Text Analysis Techniques","field":"Computer Science","cited_by":6,"is_retracted":false,"has_abstract":true,"ca_institutions":"Western University","funders":"Compute Canada","keywords":"Archetype; Jargon; Social media; Vocabulary; Computer science; Set (abstract data type); Relevance (law); Information retrieval; Slang; Data science; World Wide Web; Linguistics","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":[0.002275078,0.0007137834,0.0008078072,0.00468005,0.001059279,0.002640267,0.001771363,0.001355182,0.002306438],"category_scores_gemma":[0.01234345,0.0005507678,0.001799692,0.003727061,0.001138036,0.005985128,0.001969588,0.001304146,0.000966904],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001617694,"about_ca_system_score_gemma":0.001908765,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01259043,"about_ca_topic_score_gemma":0.01581991,"domain_scores_codex":[0.9975154,0.001102866,0.0002287742,0.0005396691,0.000492138,0.0001210867],"domain_scores_gemma":[0.9928229,0.005171692,0.0005679095,0.0007466921,0.0005581718,0.000132629],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000579558,0.0004034462,0.01945198,0.001208845,0.0003587368,0.0007533806,0.005794315,0.4253913,0.01212168,0.2885569,0.009491614,0.2358882],"study_design_scores_gemma":[0.00001234336,0.0000259993,0.0005544587,0.00002134496,0.00001809753,0.0001182894,0.0002350216,0.9444551,0.001149194,0.05045179,0.002941486,0.00001682524],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.06349514,0.0007718918,0.9242205,0.001256618,0.00005018491,0.0003558539,0.002218334,0.00111798,0.006513412],"genre_scores_gemma":[0.5707889,0.0008877563,0.4179226,0.0002449715,0.00008558288,0.0008183922,0.003452684,0.0001934557,0.005605661],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01259043,"threshold_uncertainty_score":0.02503431,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1172107362200819,"score_gpt":0.3397729561917595,"score_spread":0.2225622199716776,"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."}}