Feature Extraction Tool for Email Header Using Bidirectional Encoder Representation from Transformers (BERT) Model Constructs

Abstract

In the ever-evolving landscape of communication and information retrieval, the effective extraction of key features from emails has become a concern, especially in the field of email forensics. The current email extraction tools lack standardization in the techniques or algorithms in the extraction. Further into the privacy issue is the fear that personal information may leak if not properly handled. In this paper, an advanced email feature extraction tool using BERT  model constructs are proposed. The primary objectives include designing an email header extraction tool using an object-oriented approach, developing an email header extraction tool to extract metadata through comprehensive analysis, implementing and testing the proposed tool in terms of system functionality and user acceptance. A Natural Language Processing algorithm using the BERT model constructs the backbone of the analysis approach. The findings underscore the tool's potential to enhance information retrieval and analysis processes in various domains. In conclusion, the Feature Email Extraction Tool offers a valuable contribution to the field, providing a robust algorithm for extracting features from emails, navigating to the use of email forensics and cybersecurity.