{"id":"W162289556","doi":"","title":"Mining Multiple Web Sources Using Non-Deterministic Finite State Automata","year":2012,"lang":"en","type":"article","venue":"Scholarship at UWindsor (University of Windsor)","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Windsor","funders":"","keywords":"Computer science; Information extraction; Web page; Metadata; Information retrieval; Deterministic finite automaton; Automaton; Data extraction; Finite-state machine; Object (grammar); Data mining; Theoretical computer science; Algorithm; World Wide Web; 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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.001109903,0.0003019534,0.0004659916,0.0005359149,0.0008054092,0.0001478169,0.001938258,0.0001603348,0.00008852581],"category_scores_gemma":[0.0002351882,0.0003633702,0.0002523637,0.0009328448,0.000206229,0.002952432,0.00133949,0.0003206786,0.0002117584],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001105616,"about_ca_system_score_gemma":0.0001525241,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001266345,"about_ca_topic_score_gemma":0.00008586726,"domain_scores_codex":[0.9975072,0.0002245875,0.0002924693,0.0005947583,0.0005951006,0.0007859547],"domain_scores_gemma":[0.9975266,0.0003956787,0.0004093672,0.001129934,0.0001269602,0.0004113971],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001186141,0.0004533504,0.9328324,0.0001590678,0.0004077727,0.0002528287,0.01894391,0.001594624,0.03450464,0.000121765,0.0004241675,0.01018685],"study_design_scores_gemma":[0.004111654,0.00027568,0.3903871,0.0004776847,0.0004838667,0.0001626299,0.004060303,0.5832225,0.006062801,0.0001433603,0.008752167,0.001860269],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9746787,0.00009305518,0.02406461,0.0001290865,0.0002474874,0.0000991365,0.00008156018,0.0001474423,0.0004589224],"genre_scores_gemma":[0.9578022,0.00001200526,0.0410324,0.00006743267,0.00007317579,2.53391e-7,0.0000289342,0.00002232978,0.0009613109],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5816278,"threshold_uncertainty_score":0.9998818,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03266818413818435,"score_gpt":0.2337784940834311,"score_spread":0.2011103099452468,"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."}}