230AUTOMATED PHENOTYPING FOR DISEASE IDENTIFICATION: INTEGRATING NLP AND MACHINE LEARNING WITH SEMANTIC KNOWLEDGE IN ELECTRONIC CLINICAL NOTESEfforts are underway to explore high throughput methods for identifying patients with specific phenotypes, necessitating a standardized approach to patient identification. To automate this process using Clinic's electronic clinical notes, we employed a combination of natural language processing (NLP), machine learning, and ontology techniques. SNOMED semantic knowledge was integrated to aid in patient identification. Specifically, support vector machine (SVM) algorithms were utilized to extract SNOMED concept units from individuals both with and without type 2 diabetes mellitus (T2DM). Performance evaluation was conducted by calculating F-scores, with all concept units serving as features for both groups. The approach yielded F-scores exceeding 0.950, indicating robust performance. Patients could be classified as having a disease or syndrome based on semantic types, and even coarse concepts proved effective in detecting type 2 diabetesDeenadayalan@gmail.comResearch922019Patient identification, Natural language processing (NLP), Machine learning, Ontology, Type 2 diabetes mellitus (T2DM).Dr. DeenadayalanAssociate Professor, Department of Community Medicine, Sri Lakshmi Narayana Institute of Medical Sciences & Hospital, Osudu, Puducherry - 605502, India