{"id":"W4413799866","doi":"10.20944/preprints202508.1957.v1","title":"Text Mining and Machine Learning-Based Data Extraction for Environmental Pollutants (PPDs and PPDs-Q): Machine Learning-Enabled Global Aqueous Concentration Compilation and Potential Biotic Risk Implications","year":2025,"lang":"en","type":"preprint","venue":"Preprints.org","topic":"Water Quality and Pollution Assessment","field":"Environmental Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Jinan University","keywords":"Benthic zone; Artificial intelligence; Machine learning; Environmental science; Environmental chemistry; Chemistry; Geography; Ecology; Biology; Computer science","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.004112251,0.002032596,0.001250304,0.01837658,0.000932334,0.002488946,0.001256435,0.0008824319,0.003994338],"category_scores_gemma":[0.009925157,0.0004242692,0.002279188,0.01145658,0.0005639196,0.003143782,0.002605029,0.001004601,0.002873057],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001144935,"about_ca_system_score_gemma":0.005310046,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006730661,"about_ca_topic_score_gemma":0.0119567,"domain_scores_codex":[0.9969747,0.0005656751,0.0007520511,0.0007800724,0.0007737505,0.0001537451],"domain_scores_gemma":[0.9934112,0.003338483,0.001069547,0.0006316411,0.001343537,0.0002054597],"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":[0.00040792,0.0005156991,0.0291249,0.0176567,0.0008144354,0.001997493,0.001007525,0.00981559,0.03951033,0.003585953,0.07073863,0.8248248],"study_design_scores_gemma":[0.0002773592,0.000768332,0.09302625,0.004541192,0.002135376,0.002275437,0.002804666,0.1722169,0.1646442,0.03188078,0.524888,0.0005415638],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"empirical","genre_scores_codex":[0.08694212,0.01244248,0.2989972,0.005574251,0.0007725181,0.003048011,0.5311874,0.04515509,0.015881],"genre_scores_gemma":[0.13008,0.006970938,0.5297408,0.0007581371,0.0002425043,0.002236407,0.3232062,0.0008290941,0.005935825],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01837658,"threshold_uncertainty_score":0.02174789,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08713153731988729,"score_gpt":0.3439213031672546,"score_spread":0.2567897658473672,"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."}}