{"id":"W2170688652","doi":"10.1007/s00778-007-0067-9","title":"Applications of corpus-based semantic similarity and word segmentation to database schema matching","year":2007,"lang":"en","type":"article","venue":"The VLDB Journal","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":26,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Ottawa","funders":"","keywords":"Schema matching; Computer science; Semantic similarity; Artificial intelligence; Schema (genetic algorithms); Star schema; Natural language processing; Information retrieval; Segmentation; Pattern recognition (psychology); Database schema; Data mining; Data integration","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":[0.004904095,0.000542073,0.001111243,0.01069785,0.001374315,0.004032878,0.001548591,0.001018202,0.00392179],"category_scores_gemma":[0.03056155,0.0004899811,0.000765036,0.01182292,0.001463695,0.005064588,0.002082872,0.0008781508,0.000922015],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001422218,"about_ca_system_score_gemma":0.002580957,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006346751,"about_ca_topic_score_gemma":0.006656075,"domain_scores_codex":[0.9953908,0.001990187,0.0006650296,0.0006860617,0.001145322,0.0001225396],"domain_scores_gemma":[0.9822104,0.01159737,0.0007956791,0.002304488,0.002822255,0.0002697332],"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":[0.0006697194,0.0004790099,0.009518186,0.0007994192,0.0003343703,0.0003659847,0.00244703,0.03078659,0.02509066,0.08084552,0.00869566,0.8399679],"study_design_scores_gemma":[0.000188009,0.0002686151,0.006707333,0.0002107029,0.0002709236,0.0009488639,0.002455052,0.7357318,0.03135468,0.1949406,0.02678681,0.0001365485],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.08773176,0.001642928,0.8973171,0.00069617,0.000191193,0.000377973,0.001284557,0.003974736,0.006783556],"genre_scores_gemma":[0.3245328,0.0007950961,0.669481,0.0001539444,0.0001072612,0.000257093,0.002998144,0.0005318079,0.001142874],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01069785,"threshold_uncertainty_score":0.02593565,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1166235600128447,"score_gpt":0.4294959857511475,"score_spread":0.3128724257383028,"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."}}