Robust and Efficient Anti-Phishing Techniques.

摘要

The OSD needs high-accuracy and low-latency automatic identification and mitigation techniques to detect and stop phishing attacks. This project researched and developed socio-linguistic indicators that can be used to support more fine-grained, accurate detection of phishing emails. The indicators address several different types of phishing emails, including social malware emails, and demonstrate the feasibility and desirability of adding socio-linguistic attributes to phishing detection signatures. The project focused on : feasibility study of socio-linguistic attributes as indicators of phishing; development of socio-linguistic features for detection of phishing emails; development of a prototype for collecting socio-linguistic features from phishing emails.