Ruofan Liu

Ruofan Liu


2022
Nationality: China
Faculty and Department: Science , Statistics & Applied Probability
Year of Admission: 2020
Undergraduate University and Country: National University of Singapore (NUS) , Singapore
Thesis Advisor: Prof Dong Jin Song
Research: Phishing detection , Computer vision
https://sg.linkedin.com/in/ruofanliu

Why did you choose to do a PhD?

1. I enjoy exploring new knowledge, working with insightful people
2. My prior internship experiences make me realize that doing repetitive work is not what I want. I would like to involve in more challenging and novel projects.


Why did you choose to do graduate education at NUS? If you received offers from other universities, why did you pick NUS?

1. My undergraduate study is in NUS
2. I like the research atmosphere in my supervisor’s lab


Briefly share about your research or thesis (i.e. dissertation topic for Masters by Coursework students).

Phishing attack is one of the most common cyber attacks, causing huge financial losses every year. Prior approaches on phishing detection achieve good performances on experimental datasets. However, they are either not generalizable to wild webpages, or unable to explain its model decisions. Our research is aiming at developing a systematic phishing detection framework to: 1) achieve high precision and recall in field study 2) provide user-friendly explanations with CV techniques 3) actively adapt to human feedback.


If you have won any academic prize/competition or been invited to speak at an international conference—share what it is, its significance, and how you worked towards achieving it.

– NUS SoC Research Achievement Award in 2021/2022 Sem 1
– Temporality Spatialization: A Scalable and Faithful Time-Travelling Visualization for Deep Classifier Training. IJCAI 2022
– Inferring Phishing Intention via Webpage Appearance and Dynamics: A Deep Vision Based Approach (USENIX Security 2022)
– DeepVisualInsight: Time-Travelling Visualization for Spatio-Temporal Causality of Deep Classification Training. AAAI 2022
– Phishpedia: A Hybrid Deep Learning Based Approach to Visually Identify Phishing Webpages (USENIX Security 2021)