{"id":"W4246750514","doi":"10.1109/wi.2004.10071","title":"A Weighted Freshness Metric for Maintaining Search Engine Local Repository","year":2005,"lang":"en","type":"article","venue":"IEEE/WIC/ACM International Conference on Web Intelligence (WI'04)","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University","funders":"","keywords":"Web crawler; Computer science; Web search engine; Web page; Metric (unit); World Wide Web; Information retrieval; Search engine; Database; Web search query; Engineering","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":[0.007595615,0.001209282,0.001676851,0.007287832,0.001125425,0.003178595,0.003021741,0.00166742,0.001193304],"category_scores_gemma":[0.05717377,0.0006701359,0.0006598089,0.006049778,0.001418459,0.01187127,0.00271859,0.001387998,0.0004691187],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002349899,"about_ca_system_score_gemma":0.001790837,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002032214,"about_ca_topic_score_gemma":0.001826811,"domain_scores_codex":[0.989486,0.002059309,0.00206159,0.001633212,0.004138199,0.0006216003],"domain_scores_gemma":[0.9528996,0.01627982,0.008067179,0.009507083,0.01063993,0.002606421],"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":[0.00176694,0.0006525946,0.05382071,0.001211119,0.0006493515,0.0005761012,0.001188976,0.2134259,0.06090746,0.07878655,0.01102593,0.5759884],"study_design_scores_gemma":[0.0001100292,0.001612528,0.01870493,0.0001046538,0.0003443077,0.001428698,0.0003981417,0.8831835,0.03775228,0.03955235,0.01647544,0.0003331515],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.172333,0.002863503,0.8163314,0.0006410214,0.0002752438,0.000560433,0.001185021,0.002458274,0.0033521],"genre_scores_gemma":[0.7695138,0.0006289683,0.2258448,0.0001089244,0.0002856004,0.000394024,0.00130492,0.0003777265,0.001541344],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.007595615,"threshold_uncertainty_score":0.04016989,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06173562002561792,"score_gpt":0.3275296771444546,"score_spread":0.2657940571188366,"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."}}