{"id":"W2557589170","doi":"10.1145/3015022.3015024","title":"A Position-Based Method for the Extraction of Financial Information in PDF Documents","year":2016,"lang":"en","type":"article","venue":"","topic":"Mathematics, Computing, and Information Processing","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université du Québec à Montréal","funders":"Natural Sciences and Engineering Research Council of Canada; Mitacs","keywords":"Computer science; Heuristics; Table (database); Context (archaeology); Restructuring; Process (computing); Information extraction; Information retrieval; Position (finance); Order (exchange); General partnership; Benchmark (surveying); Data mining; Data science; Finance","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":true,"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":[0.001201738,0.001367412,0.0008012609,0.00898373,0.001136687,0.003232269,0.001147432,0.00153534,0.00779692],"category_scores_gemma":[0.007798049,0.0005169242,0.001034866,0.007889113,0.0008561997,0.002743697,0.001168471,0.001283354,0.009700888],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007391523,"about_ca_system_score_gemma":0.002557725,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004585426,"about_ca_topic_score_gemma":0.006745596,"domain_scores_codex":[0.9988006,0.0001477102,0.0001806419,0.000375836,0.0004019707,0.00009318691],"domain_scores_gemma":[0.9961042,0.001740221,0.0003792096,0.000406796,0.001254639,0.0001149129],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0002262697,0.0001389025,0.002267997,0.000686995,0.00005996499,0.0005126202,0.0005264637,0.002340663,0.02976116,0.005456354,0.02118558,0.9368371],"study_design_scores_gemma":[0.0004478462,0.0008415348,0.02771848,0.0006236144,0.0005045023,0.008327775,0.003029628,0.3037688,0.2418576,0.02758895,0.3847637,0.0005274573],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01839826,0.0012226,0.9557437,0.0005975836,0.0004042088,0.0007933028,0.005637118,0.01180996,0.005393422],"genre_scores_gemma":[0.03846283,0.0005562043,0.9491156,0.0001102445,0.0001045867,0.0002897446,0.006224315,0.000454745,0.004681732],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.00898373,"threshold_uncertainty_score":0.02608329,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01136375061371595,"score_gpt":0.2916578617110134,"score_spread":0.2802941110972974,"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."}}