{"id":"W3172351783","doi":"10.18653/v1/2021.sigdial-1.18","title":"Improving Unsupervised Dialogue Topic Segmentation with Utterance-Pair Coherence Scoring","year":2021,"lang":"en","type":"preprint","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":35,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"","keywords":"Computer science; Utterance; Coherence (philosophical gambling strategy); Artificial intelligence; Inference; Segmentation; Natural language processing; Exploit; Relevance (law); Speech recognition; Task (project management); Machine learning","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":[0.002536698,0.001600485,0.001281873,0.001469058,0.0007034728,0.0012175,0.00145322,0.001179462,0.002103527],"category_scores_gemma":[0.00891987,0.0005223829,0.0008985368,0.001075692,0.0005629999,0.002477977,0.002211963,0.002094263,0.001760282],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005982598,"about_ca_system_score_gemma":0.001425927,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004924356,"about_ca_topic_score_gemma":0.00874891,"domain_scores_codex":[0.9973449,0.001231203,0.0001145894,0.0008130993,0.0003111088,0.0001850913],"domain_scores_gemma":[0.9955771,0.002733133,0.0002546772,0.00046139,0.0007554974,0.0002181973],"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":[0.00108621,0.0006909169,0.009200456,0.000513827,0.0004315092,0.0002004449,0.002092434,0.07519821,0.0576157,0.004956938,0.01440912,0.8336042],"study_design_scores_gemma":[0.0000570868,0.0001906408,0.003634627,0.0000258901,0.0001009713,0.00007286926,0.0002760044,0.9779288,0.009946511,0.005122358,0.002606911,0.00003731461],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1032675,0.00166931,0.883767,0.0003239758,0.0001340736,0.0002102305,0.0005563956,0.007346506,0.002724925],"genre_scores_gemma":[0.7587633,0.0003603695,0.2297239,0.0002469505,0.0002911467,0.0004236948,0.00449223,0.0008682483,0.004830037],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004924356,"threshold_uncertainty_score":0.01341552,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02639506892366116,"score_gpt":0.2366871093001384,"score_spread":0.2102920403764772,"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."}}