{"id":"W2100703320","doi":"10.1109/wi.2007.132","title":"Enhancing Search Engine Quality Using Concept-based Text Retrieval","year":2007,"lang":"en","type":"article","venue":"","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":22,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Sentence; Computer science; Natural language processing; Phrase; Representation (politics); Semantics (computer science); Term (time); Artificial intelligence; Text graph; Search engine indexing; Information retrieval; Text mining","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":[0.001966164,0.001054336,0.001843092,0.006656141,0.0004295527,0.001624112,0.001421715,0.001074825,0.001720966],"category_scores_gemma":[0.00799949,0.0002123301,0.0009190956,0.00561361,0.0004614426,0.004613013,0.0009775171,0.000677877,0.0008825362],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009837666,"about_ca_system_score_gemma":0.0009660168,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003935799,"about_ca_topic_score_gemma":0.003026649,"domain_scores_codex":[0.998136,0.00035417,0.0001432775,0.0002142256,0.001062743,0.00008949476],"domain_scores_gemma":[0.9974975,0.001169034,0.0002754507,0.000226974,0.0007785735,0.00005240094],"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":[0.0009262312,0.0005433239,0.001876262,0.002025126,0.0002279686,0.0004190969,0.0005473936,0.03612744,0.08639862,0.01005395,0.01168925,0.8491653],"study_design_scores_gemma":[0.0003220194,0.0008930574,0.00507839,0.0001571894,0.0003920681,0.0009554611,0.000557454,0.9081128,0.05473152,0.01504949,0.013564,0.0001866243],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.2071739,0.01329797,0.754777,0.0008742269,0.0003070798,0.001129533,0.001653252,0.01098021,0.00980687],"genre_scores_gemma":[0.543032,0.002815252,0.4475998,0.0003378678,0.0001854503,0.0002801637,0.002779281,0.0002912871,0.002678833],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.006656141,"threshold_uncertainty_score":0.01039821,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05747550079097764,"score_gpt":0.3511540425861295,"score_spread":0.2936785417951519,"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."}}