{"id":"W3027864066","doi":"10.1145/3377939","title":"Story Forest","year":2020,"lang":"en","type":"article","venue":"ACM Transactions on Knowledge Discovery from Data","topic":"Advanced Text Analysis Techniques","field":"Computer Science","cited_by":60,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Timeline; Computer science; Event (particle physics); Information retrieval; The Internet; Novelty; World Wide Web; Cluster analysis; Set (abstract data type); News aggregator; Graph; Data science; Artificial intelligence; History","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":[0.0006922385,0.001755382,0.0006985309,0.003107492,0.001331704,0.002605887,0.002297776,0.001302518,0.07456609],"category_scores_gemma":[0.003858663,0.0006255303,0.00185235,0.002570536,0.0003868566,0.004175462,0.002283463,0.001478413,0.03810922],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006778545,"about_ca_system_score_gemma":0.001383188,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004977882,"about_ca_topic_score_gemma":0.01420694,"domain_scores_codex":[0.9991633,0.0001269304,0.00005922194,0.0003169754,0.0002345139,0.00009914594],"domain_scores_gemma":[0.9990376,0.0003215243,0.00006552265,0.0002593464,0.0002348382,0.0000811186],"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":[0.0003507535,0.0001809252,0.002870011,0.0009049827,0.0001305217,0.0004553862,0.000242131,0.01235606,0.002915271,0.02594774,0.4491531,0.5044931],"study_design_scores_gemma":[0.0001353358,0.0001163247,0.00206778,0.0002357142,0.0001019398,0.001017794,0.000442149,0.1716942,0.006384603,0.08241154,0.735324,0.00006871013],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01626917,0.003983403,0.5543593,0.002554548,0.0015691,0.002101169,0.1660955,0.08918893,0.1638788],"genre_scores_gemma":[0.09396937,0.002320024,0.4876962,0.001304079,0.0003954143,0.001380324,0.3339684,0.005827719,0.07313858],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.07456609,"threshold_uncertainty_score":0.2494484,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06664237805374586,"score_gpt":0.3114345391788544,"score_spread":0.2447921611251086,"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."}}