Removing sensitive part of a text

• Main Author: Architha, Gopinath
• Other Authors: Tay Wee Peng
• Format: Final Year Project
• Language: English
• Published: 2019
• Subjects: DRNTU::Engineering::Electrical and electronic engineering
• Online Access: http://hdl.handle.net/10356/77990
• Institution: Nanyang Technological University

With the onset of an era of digitalisation, data across many industries are now becoming digitalised. It is no surprise that the healthcare industry has moved from paper records to maintaining health records on an online portal or a system. With the vast amount of medical information in the health records, medical researchers can synthesize and find new medicine for existing diseases. They can also try to gain a more significant understanding of the underlying causes of new diseases by comparing the information across relevant medical records. With the benefits of such data sharing, it is inarguable that the same data can inevitably lead to privacy loss. Medical records contain a lot of sensitive identifiers that can easily identify the patient. From this, we can see that whenever medical records are shared for research purposes, they need to be anonymized and removed of any personal information. A combination of NLTK as well as spaCy models can be used to address this issue. With these methods, each word in the document will be allocated a meaning by the machine. Any patient identifier found, will be removed and replaced as the general PI (Patient Identifier) it refers to. This project uses Python 3.5 (64bit), NLTK 3.3.0 and spaCy. Information on the research carried out, project implementation and the results of the project are included in this report.