{"id":"W2390183027","doi":"","title":"A DOM-based Web Information Extraction","year":2007,"lang":"en","type":"article","venue":"Microcomputer applications","topic":"Advanced Computational Techniques and Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"XPath; Computer science; Document Object Model; XSLT; Relationship extraction; Path (computing); Information extraction; Data extraction; Relation (database); Information retrieval; Data mining; XML; Extraction (chemistry); World Wide Web; XML database","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0002663402,0.0001572155,0.0001092769,0.0002745596,0.0002976773,0.0001810037,0.0007473314,0.00007855497,0.000007533813],"category_scores_gemma":[8.374926e-7,0.0001684282,0.00008103858,0.0008790682,0.00003853001,0.0009742655,0.0001205502,0.0001590693,0.0002969601],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001053941,"about_ca_system_score_gemma":0.00009200536,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000004784208,"about_ca_topic_score_gemma":0.000003101401,"domain_scores_codex":[0.9987629,0.00001285414,0.0004186199,0.0003121451,0.0002146791,0.0002787833],"domain_scores_gemma":[0.9988163,0.0001355892,0.0001872113,0.0005079404,0.0002368542,0.0001161125],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.000003674737,0.0001313447,0.00003853009,0.00001272592,0.000007331056,8.381539e-7,0.00006330668,0.001738251,0.007269697,0.3007075,0.00267592,0.6873509],"study_design_scores_gemma":[0.0002536968,0.00002003664,0.0009114597,0.000006545216,0.000003779265,0.00002453075,0.000005380286,0.04826345,0.008736565,0.0165911,0.9249701,0.0002133199],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0002950683,0.0000225227,0.9936917,0.001485518,0.00002574679,0.0007642598,0.000008128281,0.00092318,0.002783892],"genre_scores_gemma":[0.1022635,0.000004173602,0.8950247,0.001998021,0.0001262709,0.0004675107,0.00006978232,0.000008940646,0.00003709538],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9222942,"threshold_uncertainty_score":0.6868303,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.006524121685157637,"score_gpt":0.280486153589012,"score_spread":0.2739620319038544,"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."}}