UMBC faculty receive NSF award for audio deepfake detection

UMBC professors Vandana Janeja, Anita Komlodi, Christine Mallinson, and Sanjay Purushotham received an NSF award to develop a prototype Community Infrastructure to Strengthen AI for Audio Deepfake analysis (CISAAD).  Deepfakes, or AI-generated content, are widely recognized as a major societal concern and challenge and audio deepfakes are increasingly being used to spread disinformation on the web and social media and for identity theft and financial scams.

The CISAAD project builds on a 2022 NSF EAGER grant to Drs. Janeja and Mallinson that studied and evaluated listener perceptions of audio deepfakes and novel interdisciplinary methods for detection. The new effort has three main goals:

  • Helping to solve the challenges of limited data availability for audio deepfakes and human-augmented data about them through open datasets shared by the community

  • Enabling  both single and multi-speaker deepfake analysis across various use cases 

  • Addressing the ethical, social, and political challenges associated with deploying deepfake technology developed from open-sourced community data

Their work will inform our understanding of mis/dis-information as a significant societal concern and challenge. It will also help understand the opportunities for content generation in positive applications such as voice restoration and smart and connected community research. 

With an interdisciplinary team across AI, linguistics, cyber infrastructure, and human-centered computing, the project develops an innovative infrastructure for expanding research informed by types of audio deepfakes. Their research and dissemination efforts will expand formal and informal learning in AI and STEM fields related to cybersecurity analytics at the intersection of technology, language, behavior, and society. The principles developed through this project will expand to multiple types of deepfakes and support media and communications experts working to address challenges related to information integrity.

CISAAD will be developed as a prototype community resource. It will include a deep fake data catalog and repository for English audio data, tools, models for deep fake audio analysis use cases, and educational material.

UMBC Center for AI ]]> UMBC professors Vandana Janeja, Anita Komlodi, Christine Mallinson, and Sanjay Purushotham received an NSF award to develop a prototype Community Infrastructure to Strengthen AI for Audio Deepfake... https://cisaad.umbc.edu/ https://dev.my.umbc.edu/api/v0/pixel/news/143016/guest@my.umbc.edu/c314aba3cb0c119c5a04508c0c439113/api/pixel ai audio cybersecurity deepfake speech UMBC AI https://dev.my.umbc.edu/groups/umbc-ai https://assets4-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xsmall.png?1691095779 https://assets2-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/original.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xxlarge.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xlarge.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/large.png?1691095779 https://assets3-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/medium.png?1691095779 https://assets3-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/small.png?1691095779 https://assets4-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xsmall.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xxsmall.png?1691095779 UMBC AI 1 0 true Thu, 01 Aug 2024 18:31:41 -0400 Talk: Tensor Decomposition for Cybersecurity, 12-1pm ET 3/29 Extracting hidden patterns from cybersecurity datasets
The UMBC Cyber Defense Lab presents

Tensor Decomposition Methods for Cybersecurity


Maksim E. Eren
Los Alamos National Laboratory

12–1pm ET, Friday, March 29, 2024, via WebEx

Tensor decomposition is a powerful unsupervised machine learning method used to extract hidden patterns from large datasets. This presentation aims to illuminate the extensive applications and capabilities of tensors within the realm of cybersecurity. We offer a comprehensive overview by encapsulating a diverse array of capabilities, showcasing the cutting-edge employment of tensors in the detection of network and power grid anomalies, identification of SPAM e-mails, mitigation of credit card fraud, and detection of malware. Additionally, we delve into the utility of tensors for classifying malware families, pinpointing novel forms of malware, analyzing user behavior, and utilizing tensors for data privacy through federated learning techniques.

Maksim E. Eren is an early career scientist in A-4, Los Alamos National Laboratory (LANL) Advance Research in Cyber Systems division. He graduated Summa Cum Laude with a Computer Science Bachelor’s at University of Maryland Baltimore County (UMBC) in 2020 and Master’s in 2022. He is currently pursuing his Ph.D. in the UMBC DREAM Lab, and he is a Scholarship for Service CyberCorps alumnus. His interdisciplinary research interests lie at the intersection of machine learning and cybersecurity, with a concentration in tensor decomposition. His tensor decomposition-based research projects include large-scale malware detection and characterization, cyber anomaly detection, data privacy, text mining, knowledge graphs, and high-performance computing. Maksim has developed and published state-of-the-art solutions in anomaly detection and malware characterization. He has also worked on various other machine learning research projects such as detecting malicious hidden code, adversarial analysis of malware classifiers, and federated learning. At LANL, Maksim was a member of the 2021 R&D 100 winning project SmartTensors, where he has released a fast tensor decomposition and anomaly detection software, contributed to the design and development of various other tensor decomposition libraries, and developed state-of-the-art text mining tools. Maksim is currently the lead for the Cyber Science Research Program (CSRP), a cybersecurity research internship at LANL.
 
Support for this event was provided in part by the National Science Foundation under SFS grant DGE-1753681.


 

 


 

 

]]> The UMBC Cyber Defense Lab presents Tensor Decomposition Methods for Cybersecurity Maksim E. Eren Los Alamos National Laboratory 12–1pm ET, Friday, March 29, 2024, via WebEx Tensor... https://ai.umbc.edu/ https://dev.my.umbc.edu/api/v0/pixel/news/140264/guest@my.umbc.edu/3de18d1b1156b5984b0bb32b31b41544/api/pixel ai cybersecurity machine-learning tensor-decomposition UMBC AI https://dev.my.umbc.edu/groups/umbc-ai https://assets4-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xsmall.png?1691095779 https://assets2-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/original.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xxlarge.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xlarge.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/large.png?1691095779 https://assets3-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/medium.png?1691095779 https://assets3-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/small.png?1691095779 https://assets4-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xsmall.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xxsmall.png?1691095779 UMBC AI Center https://assets3-dev.my.umbc.edu/system/shared/thumbnails/news/000/140/264/e44b6263328ff2868fdaa07d047fda22/xxlarge.jpg?1711567271 https://assets4-dev.my.umbc.edu/system/shared/thumbnails/news/000/140/264/e44b6263328ff2868fdaa07d047fda22/xlarge.jpg?1711567271 https://assets3-dev.my.umbc.edu/system/shared/thumbnails/news/000/140/264/e44b6263328ff2868fdaa07d047fda22/large.jpg?1711567271 https://assets4-dev.my.umbc.edu/system/shared/thumbnails/news/000/140/264/e44b6263328ff2868fdaa07d047fda22/medium.jpg?1711567271 https://assets4-dev.my.umbc.edu/system/shared/thumbnails/news/000/140/264/e44b6263328ff2868fdaa07d047fda22/small.jpg?1711567271 https://assets1-dev.my.umbc.edu/system/shared/thumbnails/news/000/140/264/e44b6263328ff2868fdaa07d047fda22/xsmall.jpg?1711567271 https://assets2-dev.my.umbc.edu/system/shared/thumbnails/news/000/140/264/e44b6263328ff2868fdaa07d047fda22/xxsmall.jpg?1711567271 0 0 true Wed, 27 Mar 2024 15:27:05 -0400 Professors Mallinson & Janeja on detecting audio deep fakes Their approach combines people, linguistics, and technology

UMBC professors Christine Mallinson and Vandana Janeja have a new article on the Maryland Matters site for news about Maryland government and politics on Bringing people and technology together to combat the threat of deep fakes. They briefly describe how they are working with a team of students to develop better ways to detect audio deep fakes with the support of an NSF grant. 

Audio deep fakes involves using AI algorithms to generate or manipulate audio recordings so that they sound like a specific person speaking or singing, typically without their consent. It poses a serious problem because it can be used to create convincing fake content for fraud, misinformation, or harassment, undermining trust in communications and media.

Their approach to detecting these deep fakes takes into account human behavioral perspectives by encouraging collaborative research across sociolinguistics, human-centered analytics, and data science.

]]> UMBC professors Christine Mallinson and Vandana Janeja have a new article on the Maryland Matters site for news about Maryland government and politics on Bringing people and technology together to... https://umbc.edu/stories/can-you-catch-a-deepfake/ https://dev.my.umbc.edu/api/v0/pixel/news/140160/guest@my.umbc.edu/06aef9d6d91e5a76f741a59b4524bb74/api/pixel ai audio-deepfakes cybersecurity linguistics UMBC AI https://dev.my.umbc.edu/groups/umbc-ai https://assets4-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xsmall.png?1691095779 https://assets2-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/original.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xxlarge.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xlarge.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/large.png?1691095779 https://assets3-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/medium.png?1691095779 https://assets3-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/small.png?1691095779 https://assets4-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xsmall.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xxsmall.png?1691095779 UMBC AI 1 0 true Mon, 25 Mar 2024 22:15:32 -0400 Online course on applying AI and HPC to cybersecurity Apply by March 30, starts on May 20, 2024

UMBC Professor Houbing Herbert Song, in collaboration with Dr. Jiawei Yuan of UMass Dartmouth, will offer an online course on applying AI and high-performance computing knowledge and skills to solve cybersecurity challenges.

The 12-week course starts on May 20, 2024. It consists of instruction in AI, machine learning, and cybersecurity, supported by teaching assistants. It is followed by faculty-guided research projects in multidisciplinary teams of participants from each area.

The pilot course is available to junior or senior undergraduate students, graduate students, postdoctoral fellows, or early-career faculty and researchers in related disciplines in a US institute at the time of participation. Those interested in participating can apply online by March 31, 2024. The course is free, and participants who successfully complete the training course will receive a $1,000 stipend.

The development of this pilot course is supported by collaborative NSF grants OAC-2309760 and OAC-2229976. 

]]> UMBC Professor Houbing Herbert Song, in collaboration with Dr. Jiawei Yuan of UMass Dartmouth, will offer an online course on applying AI and high-performance computing knowledge and skills to... http://www.cis.umassd.edu/~jyuan/cybertraining/index.html https://dev.my.umbc.edu/api/v0/pixel/news/139630/guest@my.umbc.edu/0af1ff2537568110150ef98530769bf5/api/pixel ai cybersecurity online-course UMBC AI https://dev.my.umbc.edu/groups/umbc-ai https://assets4-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xsmall.png?1691095779 https://assets2-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/original.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xxlarge.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xlarge.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/large.png?1691095779 https://assets3-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/medium.png?1691095779 https://assets3-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/small.png?1691095779 https://assets4-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xsmall.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xxsmall.png?1691095779 UMBC AI 0 0 true Tue, 05 Mar 2024 14:29:22 -0500 Talk: GotCHA: CAPTCHAs for an AI Era, 12-1 ET Friday Feb 23 Creating CAPCHAs that challenge AI systems
UMBC Cyber Defense Lab presents

GotCHA: CAPTCHAs for an AI Era

Mario Yaksetig
Cryptographer, GotCHA Labs

12-1pm ET, Friday, Feb. 23, 2024 via WebEx

In the rapidly evolving digital landscape, CAPTCHAs have transitioned from simple tests of humanity to sophisticated challenges that aim to differentiate humans and bots. Presently, CAPTCHAs are a fundamental part of securing the Internet. Unexpectedly, ChatGPT can already solve many real-world CAPTCHAs. Therefore, these AI developments result in an unforeseen problem where AIs are almost indistinguishable from real humans on the Internet. The need for advanced CAPTCHAs that can effectively thwart automated attacks has never been more critical. GotCHA is exploring short games that enhance security and contribute to the overall user experience, turning a traditionally tedious task into an engaging interaction. GotCHA's team includes a former developer from the famous Crash Bandicoot video game and a storyboard artist connected with the popular show Rick & Morty. This talk will highlight the significance of the need for new CAPTCHAs with the latest AI advancements to stay ahead of malicious bots.

Mario Yaksetig is an applied cryptographer who thrives at the intersection of privacy and the intricate world of cryptographic protocols. Mario started his cryptography journey at UMBC under Professor Alan Sherman, notable student of cryptography icon Ron Rivest. He then worked alongside David Chaum, the creator of digital cash, blind signatures, and mix networks. While working with Chaum, Mario has created multiple state-of-the-art cryptographic protocols. Recently, Vitalik Buterin, one of the creators of Ethereum, has cited some of his work on ZK-rollups. Mario's latest paper exposes a sublinear forgeability attack on a post-quantum signature scheme created by Vitalik Buterin. Email: myaksetig@gmail.com

Host: Alan T. Sherman, sherman@umbc.edu, Support for this event was provided in part by the National Science Foundation under SFS grant DGE-1753681.  The UMBC Cyber Defense Lab meets biweekly Fridays 12-1pm.  All meetings are open to the public. Upcoming CDL meetings: March 8, Cyrus Bonyadi, Metametaphysical Ontologies for Consensus; March 29, Maksim Eren, Tensor Decomposition Methods for Cybersecurity; April 12, Anupam Joshi; April 26, Dan Ragsdale, National Cybersecurity Policy; May 10, Enis Golaszewski, Automatically Binding Cryptographic Context to Messages Using Formal Methods

]]> UMBC Cyber Defense Lab presents GotCHA: CAPTCHAs for an AI Era Mario Yaksetig Cryptographer, GotCHA Labs 12-1pm ET, Friday, Feb. 23, 2024 via WebEx In the rapidly evolving digital... https://cisa.umbc.edu/ https://dev.my.umbc.edu/api/v0/pixel/news/139103/guest@my.umbc.edu/3a3aea3b772824dc679738ca88400118/api/pixel ai captcha cybersecurity UMBC AI https://dev.my.umbc.edu/groups/umbc-ai https://assets4-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xsmall.png?1691095779 https://assets2-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/original.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xxlarge.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xlarge.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/large.png?1691095779 https://assets3-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/medium.png?1691095779 https://assets3-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/small.png?1691095779 https://assets4-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xsmall.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xxsmall.png?1691095779 UMBC Cyber Defense Lab https://assets4-dev.my.umbc.edu/system/shared/thumbnails/news/000/139/103/1df1abbfe5282e744093088135b9570b/xxlarge.jpg?1708362341 https://assets3-dev.my.umbc.edu/system/shared/thumbnails/news/000/139/103/1df1abbfe5282e744093088135b9570b/xlarge.jpg?1708362341 https://assets4-dev.my.umbc.edu/system/shared/thumbnails/news/000/139/103/1df1abbfe5282e744093088135b9570b/large.jpg?1708362341 https://assets4-dev.my.umbc.edu/system/shared/thumbnails/news/000/139/103/1df1abbfe5282e744093088135b9570b/medium.jpg?1708362341 https://assets4-dev.my.umbc.edu/system/shared/thumbnails/news/000/139/103/1df1abbfe5282e744093088135b9570b/small.jpg?1708362341 https://assets1-dev.my.umbc.edu/system/shared/thumbnails/news/000/139/103/1df1abbfe5282e744093088135b9570b/xsmall.jpg?1708362341 https://assets4-dev.my.umbc.edu/system/shared/thumbnails/news/000/139/103/1df1abbfe5282e744093088135b9570b/xxsmall.jpg?1708362341 0 0 true Mon, 19 Feb 2024 12:18:15 -0500 AI Lunchbox: Ethical Hacking & Cybersecurity with HackerGPT Free online session, 12-1pm ET Thursday Feb 22, 2024
The next event in the UMBC Training Centers AI Lunchbox series is on Ethical Hacking and Cybersecurity with HackerGPT.  It will be held online from 12-1:00pm ET on Thursday, February 22, 2024.

Join us for this engaging demonstration of HackerGPT an advanced artificial intelligence program designed for ethical hacking and cybersecurity tasks. Nick Anthis will show how HackerGPT can be used to identify vulnerabilities, simulate cyber attacks, and suggest defenses in digital systems.

Register here to get a link to the free Zoom session.

]]> The next event in the UMBC Training Centers AI Lunchbox series is on Ethical Hacking and Cybersecurity with HackerGPT.  It will be held online from 12-1:00pm ET on Thursday, February 22, 2024.... https://c4a.ai/ https://dev.my.umbc.edu/api/v0/pixel/news/139080/guest@my.umbc.edu/8b060d1addc52a87eb3ad258160d0c7f/api/pixel ai cybersecurity UMBC AI https://dev.my.umbc.edu/groups/umbc-ai https://assets4-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xsmall.png?1691095779 https://assets2-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/original.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xxlarge.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xlarge.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/large.png?1691095779 https://assets3-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/medium.png?1691095779 https://assets3-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/small.png?1691095779 https://assets4-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xsmall.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xxsmall.png?1691095779 UMBC AI 1 0 true Sat, 17 Feb 2024 13:18:23 -0500 PhD defense: Knowledge Graphs & Reinforcement Learning a Hybrid Approach for Cybersecurity Tasks
Ph.D. Dissertation Defense

Knowledge Graphs and Reinforcement Learning:
a Hybrid Approach for Cybersecurity Tasks


Aritran Piplai


10:00am -12:00pm ET, Monday July 10, 2023

Dissertation Committee: Drs. Anupam Joshi (Chair/Advisor), Charles Nicholas, Chenchen Liu, Filip Perich, Joysula R. Rao, Sudip Mittal, Tim Finin

With the explosion of available data and computational power, machine learning and deep learning techniques are being increasingly used to solve problems. The domain of cybersecurity is no exception, as we have seen a significant amount of research in the recent past using data-driven machine learning approaches for different tasks.

Rule-based and supervised machine learning-based approaches are often brittle in detecting attacks, can be defeated by adversaries that adapt, and can't use the knowledge of experts. To address this problem, we propose a novel approach for cybersecurity tasks that leverages the supply of 'explicit' knowledge expressed as Knowledge Graphs, as well as the data-driven approaches of machine learning to ascertain the 'tacit' knowledge. The inspiration for this hybrid model is drawn from the manner in which security analysts work, combining their background knowledge with observed data from host and network-based sensors.

We focus on the tasks of malware detection and mitigation policy generation. We are specifically interested in the capability of synthesizing variants of a malware that can be further enhanced to detect unseen attacks. We combine the knowledge that is explicit with explorations in the 'action space' of detecting novel malware attacks as well as mitigating them. Analysts use past experiences and actions that they have taken when faced with attacks that manifest characteristics similar to the new one. The other component of their approach is exploring new `actions' that might be taken to respond to a novel malware, guided by their understanding of the domain. These steps can be considered as a 'trial and error' based approach. In this dissertation, we describe knowledge graph construction techniques from open-source text as well as several Reinforcement Learning based algorithms, guided by explicit background knowledge encoded in a knowledge graph, that best mimics the approach of security analysts. We observe that the efficiency of RL algorithms has increased by ~4% by incorporating prior knowledge. We also observe that in simulated environments RL policies are able to generate more precise mitigations with the help of prior knowledge.


]]> Ph.D. Dissertation Defense Knowledge Graphs and Reinforcement Learning: a Hybrid Approach for Cybersecurity Tasks Aritran Piplai 10:00am -12:00pm ET, Monday July 10, 2023... https://dev.my.umbc.edu/api/v0/pixel/news/134470/guest@my.umbc.edu/cee301d50a58d745031b0d8304ccb796/api/pixel cybersecurity graph knowledge UMBC AI https://dev.my.umbc.edu/groups/umbc-ai https://assets4-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xsmall.png?1691095779 https://assets2-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/original.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xxlarge.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xlarge.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/large.png?1691095779 https://assets3-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/medium.png?1691095779 https://assets3-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/small.png?1691095779 https://assets4-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xsmall.png?1691095779 https://assets1-dev.my.umbc.edu/system/shared/avatars/groups/000/002/081/cfb27ebe008c2636486089a759ea5c36/xxsmall.png?1691095779 UMBC AI 0 0 true Mon, 31 Jul 2023 12:47:46 -0400