Is Your Model Sensitive? SPEDAC: A New Resource for the Automatic Classification of Sensitive Personal Data

In recent years, there has been an exponential growth of applications, including dialogue systems, that handle sensitive personal information. This has brought to light the extremely important issue of personal data protection in virtual environments. Sensitive information detection (SID) covers different domains and languages in literature. However, if we refer to the personal data domain, the absence of a shared standard benchmark makes comparison with the state-of-the-art difficult for this task. To fill this gap, we introduce and release SPEDAC, a new annotated resource for the identification of sensitive personal data categories in the English language. SPEDAC enables the evaluation of computational models for three different SID subtasks with increasing levels of complexity. SPEDAC 1 regards binary classification, a model has to detect if a sentence contains sensitive information or not; in SPEDAC 2 we collected labeled sentences using 5 categories that relate to macro-domains of personal information; in SPEDAC 3, the labeling is fine-grained and includes 61 personal data categories. We conduct an extensive evaluation of the resource using different state-of-the-art-classifiers. The results show that SPEDAC is challenging, particularly with regard to fine-grained classification. Classifiers based on the transformer architectures achieve good results on SPEDAC 1 and 2 but have difficulties to discern among fine-grained classes in SPEDAC 3.