{"id":"W3034396897","doi":"10.1007/978-3-030-50578-3_3","title":"Structural Profiling of Web Sites in the Wild","year":2020,"lang":"en","type":"book-chapter","venue":"Lecture notes in computer science","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":false,"ca_institutions":"Université du Québec à Chicoutimi","funders":"","keywords":"Computer science; Profiling (computer programming); Information retrieval; World Wide Web; Point (geometry); Data mining; Data science; 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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0008381135,0.0003378032,0.000488747,0.0005630378,0.0001241037,0.0003488191,0.004919598,0.0001492467,0.000006970389],"category_scores_gemma":[0.0001479824,0.0002342514,0.0001301312,0.001180655,0.0004776394,0.0004378477,0.001074245,0.000747317,0.00001015583],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006282143,"about_ca_system_score_gemma":0.0003727495,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002405976,"about_ca_topic_score_gemma":0.00008286746,"domain_scores_codex":[0.9971053,0.00005832214,0.0005047229,0.001049991,0.0009009893,0.000380656],"domain_scores_gemma":[0.9978716,0.0004941434,0.0002758437,0.001188224,0.0001001031,0.00007002735],"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":[0.00001760363,0.00005557205,0.006355932,0.0002666916,0.00006952846,0.0004418856,0.009252321,0.06800844,0.004494907,0.1213097,0.000124238,0.7896032],"study_design_scores_gemma":[0.0001485144,0.0000979629,0.0002343555,0.0002051254,0.00001079013,0.00002879418,0.000001230267,0.9551056,0.001081995,0.04259581,0.000141258,0.0003485831],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.003058314,0.000428112,0.9918114,0.002144055,0.0004139879,0.0002417128,0.00001724321,0.00006699905,0.001818184],"genre_scores_gemma":[0.8121818,0.0000142029,0.1865454,0.0009892216,0.0002164373,0.000002845892,0.00001170575,0.000011907,0.0000264929],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8870971,"threshold_uncertainty_score":0.9552495,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02540800676925052,"score_gpt":0.2492081246127359,"score_spread":0.2238001178434853,"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."}}