{"id":"W4298381886","doi":"","title":"From HTML to List of Words (part 1)","year":2012,"lang":"en","type":"article","venue":"DOAJ (DOAJ: Directory of Open Access Journals)","topic":"Software Engineering and Design Patterns","field":"Social Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Western University","funders":"","keywords":"Information retrieval; Computer science; Word list; Natural language processing; World Wide Web; Index (typography)","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":[0.0008868061,0.001288737,0.0005814157,0.001929216,0.00116123,0.004619084,0.001307152,0.001249728,0.2963818],"category_scores_gemma":[0.008181957,0.0006661586,0.0005151039,0.002443186,0.0009432979,0.008076529,0.002196037,0.002845481,0.3143041],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007968454,"about_ca_system_score_gemma":0.0008704396,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001190132,"about_ca_topic_score_gemma":0.001284594,"domain_scores_codex":[0.9990576,0.0002184351,0.0001215442,0.0001623464,0.0003708251,0.00006918507],"domain_scores_gemma":[0.9967908,0.001026639,0.0001733145,0.0005459392,0.001212135,0.0002510361],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00004577961,0.00005517837,0.0002677327,0.0003044501,0.000003485342,0.0001099389,0.0008492153,0.0001919768,0.003143477,0.03228674,0.7181932,0.2445489],"study_design_scores_gemma":[0.000005624034,0.00001588306,0.0002502698,0.0001063157,0.000001727404,0.0001427471,0.0001989567,0.0002167545,0.001349271,0.01132942,0.9863701,0.00001300337],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"other","genre_gemma":"software","genre_scores_codex":[0.004375622,0.002332164,0.3438181,0.01888311,0.01250059,0.0009166633,0.01998094,0.06299832,0.5341945],"genre_scores_gemma":[0.03243056,0.003810957,0.2711827,0.01180212,0.003766228,0.0009226106,0.02140481,0.04270226,0.6119778],"genre_candidate":"software","genre_consensus":null,"teacher_disagreement_score":0.2963818,"threshold_uncertainty_score":0.991496,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3159905651206673,"score_gpt":0.574533230722056,"score_spread":0.2585426656013887,"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."}}