Computer Science New Faculty Candidate Presentations

Spring 2023

The Department of Computer Science's new faculty candidates will present their research work. All are welcome. Come and meet them.

 

Ashok Vardhan Raja 

Thursday 2/23, 9:30-10:30AM, Room (ET A331)

Ashok Vardhan Raja is a Ph.D. candidate at the University of Massachusetts Dartmouth from the Department of Computer and Information Science. He received his M.S. in Cybersecurity Engineering from the Embry-Riddle Aeronautical University in 2019 and a B.S. in Software Engineering at Rose-Hulman Institute of Technology in 2017. His research interests include Cybersecurity, UAV Security, Secure integration of AI and Cyber-physical Systems, and Adversarial Learning.

Presentation Title: Adversarial Attacks and Defenses towards the Integration of AI and UAV

Abstract: Recent years have witnessed a significant growth of UAVs in various military and civilian tasks due to their high mobility and rich sensing capabilities. This trend is further promoted by rapidly evolving AI algorithms and hardware in recent years. Although the integration of AI makes UAVs smarter and more effective, it also raises additional security and safety concerns due to potential vulnerabilities existing in the underlying AI models. The exploitation of such vulnerabilities by malicious adversaries can cause severe security and safety consequences. Therefore, it is critical to secure the integration of AI and UAV and make it robust when operating in an adversarial environment. In this talk, we will first analyze the data sensing and processing pipeline of major sensors involved in AI-enabled UAV operations to explore potential vulnerabilities. Then, we will discuss how to design defense strategies to improve the robustness of AI-enabled UAV operations. Two widely adopted AI-enabled UAV applications, including AI-enabled UAV infrastructure inspection and AI-enabled anomaly detection, will be leveraged as case studies in this talk.

 

Bhupendra Acharya

Friday 2/24, 9:30-10:30AM, Room (ET A210)

Dr. Bhupendra Acharya graduated with Doctoral Degree from the University of New Orleans (2018-2022). His research focus is on web and network security. In particular, his areas of interest lie in conducting hands-on security and privacy measurements related to web security crawlers, advertisement ecosystems, browser fingerprinting attacks, cryptocurrency scams, and other in-the-wild social engineering attacks. He is also interested in developing robust defenses against such attack vectors. Prior to joining as a Ph.D. student at UNO, he worked for eight years in an industry (Amazon, Microsoft) providing software development and assurances.

Presentation Title: Analyzing the Robustness of Prevalent Social Engineering Defense Mechanisms

Abstract: Most cybersecurity attacks begin with a social engineering attack component that exploits human fallibilities. Hence, it is very important to study the prevailing defense mechanisms against such attacks. Unfortunately, not much is known about the effectiveness of these defense mechanisms. This talk attempts to fill this knowledge gap by adopting a twofold approach that conducts a holistic analysis of social engineering attacks. In the first fold, the talk focuses on phishing attacks, which remain a predominant class of social engineering attacks despite two decades of their existence. Entities such as Google and Microsoft deploy enormous Anti-Phishing Entity systems (APEs) to enable automatic and manual visits to billions of candidate phishing websites globally. We developed a novel, low-cost framework named PhishPrint to evaluate APEs. Our framework found several flaws in APEs of 22 companies which enable attackers to easily deploy evasive phishing sites that can blindside them. We revealed all these weaknesses as well as suitable remediation measures for affected entities prompting several bugs reports as well as monetary rewards. In the second fold, the talk focuses on emerging social engineering attacks and their defense mechanisms. We chose cryptocurrency scams that run rampant on social media networks such as Twitter as an example of such emerging attacks. In order to evaluate the effectiveness of Twitter’s defense mechanisms, we developed a novel system named HoneyTweets that periodically posts messages on Twitter as bait to attract social engineering attackers. Our analysis resulted in collection of thousands of ensuing attack points such as email accounts, Instagram handles and externally hosted web pages built by attackers for the purpose of accomplishing the next stages of attacks. Our work thus presents multiple evaluation frameworks which can be used for continuous evaluation of existing social engineering defenses in future.

Gaurav Panwar

Monday 2/27, 9:30-10:30AM, Room (ET A331)

Gaurav Panwar holds a Master of Science (M.S.) in Computer Science focusing in Networking and Security from New Mexico State University (NMSU) and a Bachelor of Technology in Electronics and Communication  engineering from Mahatma Gandhi Institute of Technology, Hyderabad. He is currently a final year Ph.D student in Computer Science at New Mexico State University while also working full-time as a Senior Enterprise Network Programmer with the Enterprise Network Engineering and Design team at NMSU. Gaurav’s research interests are in Networking, Systems Security, and Cryptography. His active research areas include information-centric networking and other future internet architectures, IoT networks, smart grid technologies, security and applied cryptography in distributed systems, and Blockchain-based solutions. He has also worked in the past on design, implementation, and deployment of wireless sensor networks using microcontrollers.

Presentation Title: Building Blockchains to Support Mutable Transactions to Meet Privacy Needs.

Abstract: Blockchain technology is becoming popular for its widespread applications in areas, such as healthcare, regulatory compliance and audit, record management, and Internet of Things. Immutability of data, where transactions once posted to the blockchain cannot be modified, is one of the most significant features of Blockchains, however, there are scenarios where the ability to redact/modify certain transactions containing users’ sensitive data remains desirable. For example, a global consortium of banks (R3) is currently using a blockchain platform to manage financial agreements, securities trading, etc. These transaction records could include clients’ information and potentially contain personally identifiable information. Mutable transactions will be important to meet the needs of privacy regulations, such as the General Data Protection Regulation (GDPR). While great effort has been put into adding rewritability to Blockchains, the state-of-the-art solutions have several shortcomings, such as being coarse-grained, the inability to expunge data, the absence of revocation mechanisms, lack of user anonymity, and inefficiency. In this talk, we will discuss ReTRACe, a framework for fine-grained transaction-level blockchain rewrites, that supports revocation. ReTRACe is designed by composing a novel revocable chameleon hash with ephemeral trapdoor scheme, a novel revocable fast attribute-based encryption scheme, and a dynamic group signature scheme that come together to provide security and privacy guarantees. ReTRACe enables efficient and authorized transaction rewrites in blockchains, in addition to revocability and traceability of the users updating the transactions.

 

Mira Kim

Wednesday 3/8, 9:30-10:30AM, Room (ET A331)

Mira Kim is a technical product manager at IBM Corporation where she has been leading product strategy and technical guidelines for machine-learning document processing software systems. Dr. Kim completed her Ph.D. in Computer Science at University of California - Irvine, M.S. at University of California - Los Angeles and B.S. at University of Texas - Dallas.  Her research focuses on intelligent software engineering and reinforcement learning based recommendation systems. 

Presentation Title: Treatment Recommendation System with Hierarchical-Policy Deep Reinforcement Learning

Abstract: Reinforcement Learning is centered around an incrementally optimizing policy through a loop of taking an action, receiving a reward, and updating the policy until reaching the goal. The policy is to maintain the underlying domain knowledge that is applicable to all agents in the system. Recommendation System is to suggest the most appropriate items for users, and many of the recommendation systems reveal the behavioral variability among users in the system. This becomes evident in medical treatment recommendation systems where a single treatment method/action could result in different types and levels of effectiveness. Conventional reinforcement learning is limited in handling the complexity of treatment recommendation with the behavioral variability among agents. The objective of this presentation is to introduce a reference model for treatment recommendation systems that utilizes a hierarchy of multiple policies and deep neural network representation of the policies. The reference model includes the refinement of generating recommendation, taking an action, computing rewards, and updating the policies. The presentation includes case studies of applying the reference model on facial skin problem treatment recommender and medical treatment recommender.

 

Dong Si

Thursday, 3/9, 9:30-10:30AM, Room (ET A331)

Dr. Dong Si is currently a faculty of Computing and Software Systems, an eScience affiliated professor, and the Director of the Data Analysis & Intelligent Systems (DAIS) group at the University of Washington. Over the years, Dr. Si’s research has included machine learning and artificial intelligence, computational biology, and biomedical and health informatics. Dr. Si’s work has been published and highlighted in Nature, Nature Methods, PNAS, Nature Computational Science, WIREs, BIB, eLife, etc. The software products have been adopted by people around the world. Dr. Si has secured funding from national & state government agencies and industry. He has served as editor/reviewer/panelist/organizer for many journals/conferences/programs/events. Dr. Si has supervised more than 50 students. In addition, Dr. Si is an advocate of community engagement, diversity, equity, and inclusion. He is interested in promoting the early engagement of diverse students (especially women and underrepresented students) in artificial intelligence, biomedical and health informatics, and data science fields by introducing interdisciplinary studies, and inspiring students to pursue advanced STEM education and research careers.

Presentation Title:  "Advanced Data Analysis and Intelligent Systems for Smart Health and Next Generation Biomedicine" 

Abstract: There is an overwhelming need for the integration between computing, informatics, engineering, sciences, and biomedical disciplines to produce the innovation necessary to improve the health of the nation. And the availability of new technologies and multi-modality datasets make such integration achievable. In this presentation, I will talk about several projects in my data analysis and intelligent systems (DAIS) research group. Including DeepTracer, a fully-automated AI platform for fast de novo macromolecular structure modeling based on the 2017 Nobel Prize-winning technology cryo-electron microscopy (cryo-EM). In addition, I will also discuss other projects that support the development of novel computational approaches for the fusion and analysis of multi-level and multi-scale clinical, imaging, biomedical, personal, behavioral, and social data.