{"id":"W2101251493","doi":"10.1080/10629360701306050","title":"How can structural similarity analysis help in category formation?§","year":2007,"lang":"en","type":"article","venue":"SAR and QSAR in environmental research","topic":"Computational Drug Discovery Methods","field":"Computer Science","cited_by":62,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"European Chemical Industry Council","keywords":"Categorization; Similarity (geometry); Computer science; Ranking (information retrieval); Scope (computer science); Set (abstract data type); Process (computing); Chemical similarity; Mechanism (biology); Information retrieval; Machine learning; Data mining; Structural similarity; Artificial intelligence","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":[0.005438086,0.0006520945,0.001075791,0.003447478,0.001197874,0.002725247,0.001618453,0.00155958,0.008756449],"category_scores_gemma":[0.01993433,0.0003041464,0.001420985,0.002244927,0.001888892,0.005210157,0.00194896,0.001538023,0.002652538],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009344318,"about_ca_system_score_gemma":0.00119203,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001428762,"about_ca_topic_score_gemma":0.001753615,"domain_scores_codex":[0.9974706,0.001291594,0.0001854797,0.0004555267,0.0004724903,0.0001242015],"domain_scores_gemma":[0.9905422,0.00544499,0.0008424756,0.001584083,0.001318025,0.0002681277],"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":[0.0002284435,0.0003202312,0.01531465,0.000582188,0.0001452387,0.0001652191,0.0006363148,0.03652225,0.006376978,0.153368,0.009087904,0.7772527],"study_design_scores_gemma":[0.00006359004,0.000428722,0.007805628,0.0001528534,0.00009392318,0.0004584523,0.001011641,0.2935822,0.00986379,0.6631252,0.02330974,0.0001044124],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0974138,0.001477009,0.8771931,0.004718183,0.0003488326,0.0003354255,0.0007028555,0.002471265,0.01533958],"genre_scores_gemma":[0.4557794,0.000557952,0.537825,0.000447361,0.0001707451,0.0002227437,0.001146987,0.0002297208,0.003620002],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.008756449,"threshold_uncertainty_score":0.0292933,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04607903939175301,"score_gpt":0.3595187392152965,"score_spread":0.3134396998235435,"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."}}