Research Experience ¶
- Cybercrime group, Cylab, Carnegie Mellon University (Sep 2019 - Present)
- Graduate Research Assistant, advisor: Prof. Nicolas Christin. Pittsburgh, PA, USA.
- Non-fungile token liquidity measurement: Comparing to typical DeFi coins lending and borrowing, NFT loans have a significantly higher APY/APR. We quantitatively measure this phenomenon and compare it to the existing payday loan.
- Malicious activities in Tor: Anecdotally, malicious actors test malware through Tor network before distributing them onto the regular network. We compare the traffic alert distribution captured at a Tor exit node to that of the regular network.
- Web-browsing exposure prediction: A typical blocklist-based security defense system suffers from an inherent gap between malicious content creation and detection. We develop a proactive defense system that is robust against concept drift to classify whether users’ web-browsing behavior may lead to malicious content exposure.
- Cryptocurrency market measurement: Cryptocurrency market has evolved from spot trading to complex derivative trading since 2018. We analyze one of the most popular trading markets and describe the diverse ensemble of amateur and professional traders with on-chain forensics and behavior clustering. (WWW’21)
- Vision and Science Lab, National Tsing-Hua University (Jul 2017 - Jun 2018)
- Research Assistant, advisor: Prof. Min Sun. Hsinchu, Taiwan.
- Device-aware Neural Architecture Search: Designed a multi-objective deep model architecture search framework to generate neural networks given different computational constraints. (ECCV’18)
Selected Publications ¶
Towards Understanding Cryptocurrency Derivatives: A Case Study of BitMEX
- Kyle Soska, Jin-Dong Dong, Alex Khodaverdian, Ariel Zetlin-Jones, Bryan Routledge, Nicolas Christin
- Proceedings of the Web Conference 2021, Ljubljana, Slovenia
DPP-Net: Device-aware Progressive Search for Pareto-optimal Neural Architectures
- Jin-Dong Dong, An-Chieh Cheng, Da-Cheng Juan, Wei Wei, Min Sun
- Proceedings of the European Conference on Computer Vision (ECCV) 2018, Munich, Germany