{"id":"W1494273468","doi":"10.1007/978-3-540-24840-8_5","title":"Term-Based Clustering and Summarization of Web Page Collections","year":2004,"lang":"en","type":"book-chapter","venue":"Lecture notes in computer science","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":20,"is_retracted":false,"has_abstract":false,"ca_institutions":"Dalhousie University","funders":"","keywords":"Automatic summarization; Computer science; Cluster analysis; Information retrieval; Web page; World Wide Web; Web intelligence; Task (project management); Web modeling; Artificial intelligence","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":[0.001009519,0.0007434046,0.001548607,0.009978835,0.001111046,0.002054386,0.001868657,0.0008529449,0.002599862],"category_scores_gemma":[0.004416548,0.000451987,0.001336563,0.009254218,0.0003893928,0.001661698,0.0008646692,0.0008126217,0.002780601],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007424315,"about_ca_system_score_gemma":0.001253166,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008637242,"about_ca_topic_score_gemma":0.01746238,"domain_scores_codex":[0.9989036,0.0001586692,0.0001311827,0.0002525955,0.0004245848,0.0001292826],"domain_scores_gemma":[0.9975554,0.0006629623,0.0002130882,0.0003735333,0.001068105,0.0001268971],"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":[0.0005100537,0.0002421575,0.00462392,0.0005371593,0.0002680649,0.0001631086,0.0004377119,0.01368672,0.04556272,0.00164652,0.01588754,0.9164343],"study_design_scores_gemma":[0.0001328534,0.0007851241,0.06141137,0.0002057778,0.001440115,0.001664304,0.001268841,0.7871545,0.07971615,0.02474141,0.04118778,0.0002918094],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.167441,0.00747298,0.8023871,0.0004488439,0.0003833765,0.0005657463,0.007392457,0.009327629,0.004580884],"genre_scores_gemma":[0.2688611,0.002547311,0.68405,0.00008883249,0.0005112191,0.0003397551,0.03130795,0.0009695583,0.01132428],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.009978835,"threshold_uncertainty_score":0.01717395,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01384464157001871,"score_gpt":0.2280352087547077,"score_spread":0.214190567184689,"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."}}