{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.008423096,0.0007570047,0.001254123,0.003400122,0.00136347,0.003741766,0.001717226,0.001280888,0.002309874],"category_scores_gemma":[0.05652448,0.0005609596,0.001410424,0.006118911,0.003370242,0.009865261,0.004987735,0.003496836,0.001138115],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009732803,"about_ca_system_score_gemma":0.001937619,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002053448,"about_ca_topic_score_gemma":0.001326495,"domain_scores_codex":[0.9830012,0.005697199,0.001689926,0.004465098,0.004443241,0.0007033274],"domain_scores_gemma":[0.9689516,0.0158279,0.002611613,0.007168867,0.004862701,0.0005773163],"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":[0.0005314193,0.0002328154,0.01210126,0.0007246095,0.0003545183,0.0005710213,0.002190376,0.02495646,0.01345443,0.6763448,0.005017316,0.263521],"study_design_scores_gemma":[0.000103025,0.0003826292,0.006915252,0.0001821361,0.0002088664,0.001324418,0.0009880337,0.2181762,0.007192452,0.7450852,0.01929055,0.0001512395],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05363547,0.001115407,0.9385213,0.000733191,0.0002255833,0.0001379764,0.000449594,0.0001805483,0.005000849],"genre_scores_gemma":[0.72023,0.002368819,0.2694232,0.0005731027,0.0009327341,0.0006957051,0.001889621,0.0001491243,0.003737646],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.008423096,"threshold_uncertainty_score":0.04454607,"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."}}