Showing 1–1 of 1 results for author: Larbi, I B C

Which anonymization technique is best for which NLP task? -- It depends. A Systematic Study on Clinical Text Processing

Authors: Iyadh Ben Cheikh Larbi, Aljoscha Burchardt, Roland Roller

Abstract: Clinical text processing has gained more and more attention in recent years. The access to sensitive patient data, on the other hand, is still a big challenge, as text cannot be shared without legal hurdles and without removing personal information. There are many techniques to modify or remove patient related information, each with different strengths. This paper investigates the influence of dif… ▽ More Clinical text processing has gained more and more attention in recent years. The access to sensitive patient data, on the other hand, is still a big challenge, as text cannot be shared without legal hurdles and without removing personal information. There are many techniques to modify or remove patient related information, each with different strengths. This paper investigates the influence of different anonymization techniques on the performance of ML models using multiple datasets corresponding to five different NLP tasks. Several learnings and recommendations are presented. This work confirms that particularly stronger anonymization techniques lead to a significant drop of performance. In addition to that, most of the presented techniques are not secure against a re-identification attack based on similarity search. △ Less

Submitted 1 September, 2022; originally announced September 2022.