{"id":"W2044194343","doi":"10.4018/jdwm.2012100101","title":"Towards Comparative Mining of Web Document Objects with NFA","year":2012,"lang":"en","type":"article","venue":"International Journal of Data Warehousing and Mining","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":6,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Windsor","funders":"","keywords":"Computer science; Web mining; Data Web; Web page; Web modeling; Information retrieval; Web standards; World Wide Web; Web intelligence; Static web page; Web navigation; Database","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":[0.002119965,0.0006593987,0.001033929,0.007591461,0.0009912963,0.002076074,0.001326512,0.001191798,0.001054557],"category_scores_gemma":[0.009742036,0.0005840982,0.002355622,0.004471871,0.001203348,0.002929386,0.00121062,0.0007356054,0.0005253489],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001273268,"about_ca_system_score_gemma":0.001358606,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006823195,"about_ca_topic_score_gemma":0.006116615,"domain_scores_codex":[0.9982967,0.0003474304,0.0002189699,0.000534926,0.0005102107,0.0000918649],"domain_scores_gemma":[0.9958751,0.002483474,0.0003552868,0.0005237614,0.0006861245,0.00007633213],"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":[0.0004381681,0.0003543235,0.03436643,0.0008768069,0.0002914202,0.001130078,0.001590031,0.1013671,0.04258334,0.04182286,0.001648498,0.7735308],"study_design_scores_gemma":[0.00002048558,0.0001588173,0.008322048,0.0001307962,0.0001212701,0.0008504272,0.0003891366,0.9055477,0.01814821,0.05670725,0.009559535,0.00004432174],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05184272,0.0004870047,0.9435182,0.00008934259,0.00001263193,0.0001431041,0.0005537666,0.002001055,0.001352248],"genre_scores_gemma":[0.2126353,0.000241886,0.7837634,0.00004243396,0.00001471639,0.0002651915,0.002085032,0.0001344975,0.0008174551],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007591461,"threshold_uncertainty_score":0.01356691,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07416393370134935,"score_gpt":0.3429523755542617,"score_spread":0.2687884418529124,"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."}}