{"id":"W2096565906","doi":"","title":"Alignment-Based Discriminative String Similarity","year":2007,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":54,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Discriminative model; Coreference; Artificial intelligence; Computer science; Substring; String metric; Similarity (geometry); Character (mathematics); Natural language processing; Longest common subsequence problem; String (physics); Word (group theory); Pattern recognition (psychology); Transliteration; Heuristic; String searching algorithm; Mathematics; Pattern matching; Resolution (logic); Algorithm; Set (abstract data type)","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":[0.00154235,0.0006003175,0.001530709,0.003846962,0.0006893366,0.001102409,0.001571028,0.0007734903,0.00250752],"category_scores_gemma":[0.007312349,0.0002540779,0.0005424664,0.004968217,0.0008512837,0.002222912,0.001642399,0.000862533,0.001794837],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005512426,"about_ca_system_score_gemma":0.000798322,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0008711013,"about_ca_topic_score_gemma":0.001519419,"domain_scores_codex":[0.9973628,0.0006219286,0.0001683611,0.0006997891,0.0009777171,0.0001694113],"domain_scores_gemma":[0.9960316,0.001126095,0.0005429389,0.001282823,0.000843252,0.0001732548],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0004615589,0.000388522,0.01324662,0.0004186064,0.0001552072,0.0001825583,0.000245671,0.0347811,0.08358547,0.0316047,0.007315861,0.8276141],"study_design_scores_gemma":[0.00007156301,0.0008021127,0.01997499,0.00003487004,0.0001034421,0.001716093,0.000194624,0.8387704,0.07313561,0.05217202,0.01290158,0.0001227065],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.07159048,0.0004947179,0.9210771,0.00008972681,0.00006303943,0.0001408528,0.0004943854,0.002436623,0.003612904],"genre_scores_gemma":[0.6587179,0.0002156448,0.3359312,0.0001368636,0.00009083121,0.0001661734,0.002211272,0.0002629427,0.002267076],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.003846962,"threshold_uncertainty_score":0.00838846,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01628738332415331,"score_gpt":0.2953723765151392,"score_spread":0.2790849931909859,"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."}}