{"id":"W2507709051","doi":"10.1145/2960811.2960813","title":"Relaxing Orthogonality Assumption in Conceptual Text Document Similarity","year":2016,"lang":"en","type":"article","venue":"","topic":"Advanced Text Analysis Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University","funders":"","keywords":"Orthogonality; Similarity (geometry); Cosine similarity; Information retrieval; Computer science; Space (punctuation); Measure (data warehouse); Similarity measure; Key (lock); Natural language processing; Artificial intelligence; Data mining; Mathematics; 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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0006069969,0.0001174041,0.0001672279,0.0001102339,0.00004780381,0.0000443692,0.0005289799,0.00006755923,0.0001546887],"category_scores_gemma":[0.0001027032,0.000078538,0.00006604454,0.0003265183,0.00008448685,0.001177026,0.0002874765,0.00009953947,0.00007754451],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001960729,"about_ca_system_score_gemma":0.00003123998,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004929285,"about_ca_topic_score_gemma":0.0002357687,"domain_scores_codex":[0.9986066,0.0001153824,0.0003087066,0.000431198,0.0002862361,0.0002519236],"domain_scores_gemma":[0.9990696,0.0001609256,0.00009026428,0.0005542345,0.00006032998,0.00006462508],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.000005794338,0.000072284,0.04029271,0.00000304484,0.00001278757,0.00001473063,0.000218107,0.00001890432,0.006933616,0.7827454,0.0005330351,0.1691496],"study_design_scores_gemma":[0.001964875,0.0002274445,0.211686,0.0001929372,0.00001824092,0.00001676123,0.0001169554,0.005723301,0.09904514,0.647534,0.03212829,0.001346045],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05214957,0.00002804103,0.9420997,0.001832974,0.00004127697,0.0001067671,5.247588e-7,0.0003810624,0.003360123],"genre_scores_gemma":[0.8826103,0.00002519874,0.1164285,0.0002740477,0.00001769064,0.00001657607,6.014037e-7,0.00000424879,0.0006228772],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8304607,"threshold_uncertainty_score":0.3202687,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01983061498926167,"score_gpt":0.2893450317022357,"score_spread":0.269514416712974,"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."}}