Multi-Task Romanian Email Classification in a Business Context

Abstract

Email classification systems are essential for handling and organizing the massive flow of communication, especially in a business context. Although many solutions exist, the lack of standardized classification categories limits their applicability. Furthermore, the lack of Romanian language business-oriented public datasets makes the development of such solutions difficult. To this end, we introduce a versatile automated email classification system based on a novel public dataset of 1447 manually annotated Romanian business-oriented emails. Our corpus is annotated with 5 token-related labels, as well as 5 sequence-related classes. We establish a strong baseline using pre-trained Transformer models for token classification and multi-task classification, achieving an F1-score of 0.752 and 0.764, respectively. We publicly release our code together with the dataset of labeled emails.

Lambda Architecture-Based Big Data System for Large-Scale Targeted Social Engineering Email Detection

“…Additionally, the proponents provided a rationale for the effective capability of the proposed model in identifying spam emails. In another recent study [17], the authors presented a highly adaptable automated email classification system that utilizes a unique publicly available dataset consisting of 1447 Romanian business-oriented emails that were manually annotated. A robust foundation was established by employing pre-trained Transformer models for token classification and multi-task classification, resulting in F1 scores of 0.752 and 0.764, respectively.…”