{"id":"W2158255858","doi":"10.5555/1937055.1937105","title":"Towards a collection-based results diversification","year":2010,"lang":"en","type":"article","venue":"","topic":"Information Retrieval and Search Behavior","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Information retrieval; Computer science; Cosine similarity; Diversification (marketing strategy); Relevance (law); Data mining; Document retrieval; Artificial intelligence; Cluster analysis","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":[0.01503329,0.00136913,0.002644573,0.006955896,0.001820602,0.004899848,0.003731665,0.002083554,0.001880968],"category_scores_gemma":[0.02458891,0.0009431092,0.001790654,0.005472209,0.002128726,0.006022152,0.006235823,0.002263202,0.001636361],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001285253,"about_ca_system_score_gemma":0.002330534,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001550192,"about_ca_topic_score_gemma":0.001680108,"domain_scores_codex":[0.9839029,0.006560265,0.0008672587,0.002127071,0.006070228,0.0004722199],"domain_scores_gemma":[0.9769596,0.0065079,0.00150336,0.008179037,0.006059057,0.0007909739],"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":[0.001158196,0.001341522,0.01065721,0.0008761437,0.0006431458,0.000372298,0.002407473,0.04086776,0.1179636,0.03992824,0.01200448,0.7717798],"study_design_scores_gemma":[0.0005973207,0.001812122,0.008215424,0.000299004,0.0007483307,0.002842611,0.001035348,0.6992797,0.1342985,0.09386121,0.05666321,0.0003471787],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.032132,0.001631255,0.9585027,0.0007859442,0.00005955667,0.0004285881,0.0001359591,0.002301246,0.004022717],"genre_scores_gemma":[0.1533194,0.0005497178,0.8412296,0.0004853066,0.0001781865,0.0003846054,0.0003844776,0.0005287941,0.002939942],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01503329,"threshold_uncertainty_score":0.07950461,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02275603600988979,"score_gpt":0.2648521647651605,"score_spread":0.2420961287552708,"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."}}