Dutch Named Entity Recognition and De-identification Methods for the Human Resource Domain
This work addresses privacy protection in HR documents, but it is incremental as it updates existing methods and tests new datasets.
The paper tackled the problem of de-identifying privacy-sensitive textual data in the human resource domain by evaluating and updating Dutch named entity recognition methods, achieving high recall rates such as 0.94 for persons and 0.82 for locations with a BERTje-based model.
The human resource (HR) domain contains various types of privacy-sensitive textual data, such as e-mail correspondence and performance appraisal. Doing research on these documents brings several challenges, one of them anonymisation. In this paper, we evaluate the current Dutch text de-identification methods for the HR domain in four steps. First, by updating one of these methods with the latest named entity recognition (NER) models. The result is that the NER model based on the CoNLL 2002 corpus in combination with the BERTje transformer give the best combination for suppressing persons (recall 0.94) and locations (recall 0.82). For suppressing gender, DEDUCE is performing best (recall 0.53). Second NER evaluation is based on both strict de-identification of entities (a person must be suppressed as a person) and third evaluation on a loose sense of de-identification (no matter what how a person is suppressed, as long it is suppressed). In the fourth and last step a new kind of NER dataset is tested for recognising job titles in texts.