Indonesian Protected Health Information Removal using Named Entity Recognition

The Electronic Health Record (EHR) is the implementation of a computer system to keep a history of patient health records. All information related to patients is stored as patients who have medical history, immunization status, sugar level, and blood pressure records. EHR also contains the patient protected health information such as patient name and patient id. However, before processing further any text processing method, the protected health information needs to be removed from the data. This research employs a combination of long short term memory neural networks and conditional random fields in order to remove protected health information within the data. We employ several evaluations criteria to discover the best model. Our final evaluation result yields 0.76 the proposed model best result.