{"id":"W2986143601","doi":"10.18653/v1/k19-1020","title":"Fully Unsupervised Crosslingual Semantic Textual Similarity Metric Based on BERT for Identifying Parallel Data","year":2019,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":20,"is_retracted":false,"has_abstract":true,"ca_institutions":"National Research Council Canada","funders":"","keywords":"Computer science; Semantic similarity; Artificial intelligence; Metric (unit); Similarity (geometry); Natural language processing; Information retrieval; Pattern recognition (psychology)","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":[0.001798283,0.000985696,0.0008163564,0.003916814,0.0007170222,0.001228393,0.001196306,0.0009406242,0.002342829],"category_scores_gemma":[0.008854412,0.0002357693,0.0007517515,0.002831555,0.0008544112,0.003415646,0.002602889,0.00116532,0.001719245],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008179013,"about_ca_system_score_gemma":0.001188098,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002989565,"about_ca_topic_score_gemma":0.005270708,"domain_scores_codex":[0.9974075,0.0007883518,0.0002654376,0.0007305536,0.000644188,0.0001640247],"domain_scores_gemma":[0.9964381,0.001012624,0.0004063887,0.0008584833,0.001133329,0.0001510479],"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":[0.001032167,0.0005073724,0.01894696,0.0008793144,0.0003054478,0.0005035138,0.001097531,0.05776237,0.0543652,0.02542922,0.01833662,0.8208343],"study_design_scores_gemma":[0.00006073693,0.0004250666,0.01406555,0.0000830379,0.0001298264,0.0008128729,0.0008203601,0.8811128,0.04305773,0.03774379,0.02156872,0.0001195261],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1698488,0.001292224,0.8122032,0.0002580219,0.0001564694,0.0002984393,0.003983625,0.005872102,0.006087089],"genre_scores_gemma":[0.6830339,0.0003179816,0.2945679,0.0001468178,0.0001027323,0.0004864597,0.01709507,0.0006350394,0.003613995],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.003916814,"threshold_uncertainty_score":0.009510338,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1050365625364869,"score_gpt":0.3355009182367941,"score_spread":0.2304643557003073,"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."}}