Year: 2020
Type: dissertation
Abstract: With the introduction of the General Data Protection Regulation (GDPR) the demand to protect personal identi able information from unnecessary exposure is growing.The objective of this thesis is to explore and construct a system that will use natural language processing and machine learning to anonymize data.This will provide further analysis opportunities and additional protection for the individual person.We implemented and compared di erent neural network architectures and frameworks.Using a bidirectional Long Short-Term Memory network together with contextualized embeddings and conditional random elds, state-of-the-art results were reached with the best F 1 score at .%.This shows the importance of contextual awareness as well as the impact of pretrained word embeddings.However, human supervision will still be needed for it to completely guarantee the anonymization of the messages.Ultimately, the resulting system ful lls its purpose of masking texts, removing the risk of unnecessary information exposure.
Authors: John Helbrink, Simon Åkesson
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W3047587672