{"id":"W2139676742","doi":"10.48550/arxiv.1307.2669","title":"Text Categorization via Similarity Search: An Efficient and Effective Novel Algorithm","year":2013,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Ottawa","funders":"","keywords":"Computer science; Similarity (geometry); Centroid; Cluster analysis; Categorization; Artificial intelligence; Outlier; Preprocessor; Metric (unit); Text categorization; Class (philosophy); Point (geometry); Similarity measure; Measure (data warehouse); Nearest neighbor search; Pattern recognition (psychology); Data mining; Mathematics; Image (mathematics)","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":[0.001380561,0.001136058,0.0020887,0.004363592,0.0009899433,0.001789956,0.002588333,0.002038543,0.004212876],"category_scores_gemma":[0.004893463,0.0004802512,0.001032563,0.005564223,0.0007982156,0.00397995,0.002199961,0.001479762,0.004687806],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008038005,"about_ca_system_score_gemma":0.001783521,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001826797,"about_ca_topic_score_gemma":0.001984697,"domain_scores_codex":[0.9972181,0.000502413,0.0002430893,0.0006209747,0.001272043,0.0001433531],"domain_scores_gemma":[0.9983296,0.0005405246,0.0001280741,0.0003149573,0.0006152509,0.00007166024],"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":[0.0002183547,0.0002977453,0.0008128636,0.0001801888,0.00007199611,0.0001120799,0.0001042252,0.01532779,0.01356591,0.009165956,0.01400434,0.9461385],"study_design_scores_gemma":[0.000196401,0.0002596717,0.0008001449,0.00003529145,0.00006238899,0.0007754479,0.0001295024,0.9257344,0.01290311,0.03725848,0.02179279,0.0000524737],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.006250342,0.000532118,0.9889832,0.0002694335,0.0001309231,0.00021589,0.000167753,0.002176321,0.001273988],"genre_scores_gemma":[0.05567879,0.0002898018,0.9386085,0.0002018928,0.0001923143,0.0003414638,0.0009634907,0.0001679293,0.003555779],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.004363592,"threshold_uncertainty_score":0.01409352,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06421189842904142,"score_gpt":0.202638992897949,"score_spread":0.1384270944689076,"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."}}