September 2023 - Gradstudents - mailman.cs.vt.edu

Jie Bu's Ph.D. defense [Feb 9 at 4 pm]
by Anuj Karpatne 02 Dec '24

02 Dec '24

Hello everyone, You are cordially invited to the final defense of PhD student Jie Bu on Thursday, Feb 9 at 4 pm. Zoom Link: https://virginiatech.zoom.us/j/5142223333 Title: Achieving More with Less: Learning Generalizable Neural Networks With Less Labeled Data and Computational Overheads Committee: Dr. Anuj Karpatne (advisor) Dr. Adrian Sandu Dr. Ismini Lourentzou Dr. Naren Ramakrishnan Dr. Anish Arora (from Ohio State University) Abstract: Recent advancements in deep learning have demonstrated its incredible ability to learn generalizable patterns and relationships automatically from data in a number of mainstream applications. However, the generalization power of deep learning methods largely comes at the costs of working with very large datasets and using highly compute-intensive models. Many applications cannot afford these costs needed to ensure generalizability of deep learning models. For instance, obtaining labeled data can be costly in scientific applications, and using large models may not be feasible in resource-constrained environments involving portable devices. This thesis aims to improve efficiency in machine learning by exploring different ways to learn generalizable neural networks that require less labeled data and computational resources. We demonstrate that using physics supervision in scientific problems can reduce the need for labeled data, thereby improving data efficiency without compromising model generalizability. Additionally, we investigate the potential of transfer learning powered by transformers in scientific applications as a promising direction for further improving data efficiency through domain adaptation. On the computational efficiency side, we present two efforts for increasing parameter efficiency of neural networks through novel architectures and structured network pruning. Thanks, -Anuj ------------------- Anuj Karpatne Assistant Professor Computer Science, Virginia Tech https://people.cs.vt.edu/karpatne/

2 6

Muskan Gupta's Thesis exam
by Sang Won Lee 02 May '24

02 May '24

Hi, You are cordially invited to Muskan Gupta's thesis exam. I've included the details below. Sang Won Lee *When/Where* 8/31/2023 (Thursday) 1 PM-3 PM The Gilbert Place Room 3211 *Committee* Sang Won Lee (Chair) Eugenia Rho Scott McCrickard *Title*: Understanding Social Media Users' Perceptions of Trigger and Content Warnings *Abstract*: Exposure to distressing content on social media poses risks to users' mental well-being, leading to the use of trigger warnings (TW) and content warnings (CW). However, there is limited insight into user preferences and the factors affecting these warnings' effectiveness. This study addressed this gap through qualitative interviews with 15 participants. Findings revealed generally positive user perceptions of warnings, which provide protection from distressing content, create safe spaces for sharing sensitive experiences, and avoid such content in public spaces. However, understanding when, what, and how to use warnings varied among users, leading to complex decision-making. Factors influencing warning effectiveness included their separation from content, length, awareness, consistency, context, severity, reasoning, creator's intent, and audience knowledge. Participants stressed the need for enhanced education on warnings and triggers in social media. The study produced design recommendations, including personalized filtering, platform guidelines, specificity, nudging creators, interventions against excessive scrolling, accessibility improvements, and community feedback mechanisms. Participants also suggested post-trigger support measures. The study uncovers challenges in balancing warnings with user engagement, specificity, and personalization while avoiding the creation of filter bubbles. While the study spanned various platforms, future research should focus on specific platforms to address limitations like recruitment bias and broad scope. Future directions include incentivizing warning use, implementing nudges and interventions, and improving post-trigger support, as well as investigating audio triggers and quantitative studies for refined TW/CW design to promote a trauma-informed social media environment.

2 2

grad seminar this Friday, Sep 22, 2:30 PM Torgersen 3100
by Calvin Ribbens 11 Oct '23

11 Oct '23

All are invited to hear short intro talks from six CS faculty members who have joined the department in the last few years: Tijay Chung, Atul Matnri, Krish Sundararajah, Yaxing Yao, Pinar Yanardag, Dimitris Nikolopoulos. This will be a great opportunity to sample the breadth of research going on in CS at Virginia Tech. The presentations will be available on zoom <https://virginiatech.zoom.us/j/89995088537?pwd=NCtqMlNBTEV0Q3QvOENVR2FXRDNa…> as well, for those who are unable to attend in person. Dr. Ribbens

1 1

two seminars of possible interest next week
by Calvin Ribbens 29 Sep '23

29 Sep '23

CS Gradstudents, We are hosting two visitors to the department next week Tuesday and Wednesday, with talks set for each day. Both will be available on Zoom. Dr. Ribbens ----- *Tuesday, October 3: Karen Sowon* *Seminar: Digital Financial Inclusion: Challenges in the Use of IDs for Mobile Money Services <https://website.cs.vt.edu/research/Seminars/Karen-Sowon.html>* *10:00 – 11:00am, 1100 Torgersen Hall* Join Seminar via Zoom: https://virginiatech.zoom.us/j/83811477522?pwd=TC84QlhSeCtQeURzbEQzbWlzbnQx… Meeting ID: 838 1147 7522 | Passcode: 285425 *Wednesday, October 4: Renato Figueiredo* *Seminar: Edge-to-Cloud Virtualization and Applications in Ecological Forecasting <https://website.cs.vt.edu/research/Seminars/Renato-Figueiredo.html>* *10:00 – 11:00am; 1100 Torgersen Hall* Join Seminar via Zoom: https://virginiatech.zoom.us/j/88639714634?pwd=WUZEY2E3OWNJd2Rad1pRTDgvdDNl… Meeting ID: 886 3971 4634 | Passcode: 491557

1 0

Fwd: NIU CS Faculty Positions
by Chris North 28 Sep '23

28 Sep '23

From a VT alum at NIU: > Begin forwarded message: > > From: Maoyuan Sun <smaoyuan(a)gmail.com> > Subject: NIU CS Positions > Date: September 28, 2023 at 9:41:21 AM EDT > To: Chris North <north(a)cs.vt.edu> > > ... > > My department has 5 openings this year. One senior position for cybersecurity, and four for broadly data related (AI, ML, data science) ones. The hiring ad can be found from the following link: > > https://cra.org/job/northern-illinois-university-assistant-professor-of-com… > > Please share this with VT grads. > > Thanks, > Maoyuan

1 0

FW: [Sanghani-faculty] Sanghani Seminar Series ft. Stéphane F. Maison, PhD, AuD, CCC-A, Sept. 28
by Sharon Kinder-Potter 26 Sep '23

26 Sep '23

Sharon Kinder-Potter Graduate Program Coordinator Virginia Tech Department of Computer Science 3210 Torgersen Hall; (MC 0106) 620 Drillfield Drive Blacksburg, VA 24061 540.231.6932/ <mailto:skpotter@vt.edu> skpotter(a)vt.edu <https://cs.vt.edu/> https://cs.vt.edu **BE COMMITTED. BE WELL** _____

1 0

CS Grad Council Meeting -- Monday 9/25 from 12:30 to 1:30 PM in Gilbert 3001
by Wesley Woo 24 Sep '23

24 Sep '23

Hi all, Just a reminder that there will be a CS Grad Council Meeting *tomorrow* (09/25) from *12:30 pm to 1:30 pm in Gilbert 3001*! This meeting will be hybrid (link <https://virginiatech.zoom.us/j/84656150447>) We will be discussing things including, but not limited to: - Creating our events list and budget for the semester - A CSGC lightning talks event - Finalizing mentorship program planning - CSGC social media presence? If you are interested in contributing to the discussion around any of these (*read as: helping us plan the free food events for the semester*), stop by *tomorrow at 12:30 to 1:30 PM in Gilbert 3001*! Hope to see everyone there! (*tomorrow, Gilbert 3001, 12:30 to 1:30 PM*) Wesley Woo CS Grad Council Secretary

1 0

Tung Dao's PhD dissertation defense scheduled at 10am on 9/29
by Na Meng 20 Sep '23

20 Sep '23

Hi all, You're invited to attend Tung's PhD dissertation defense. *Title:* Toward More Practical Fault Localization with Instant Localization and Triggering Modes: An Empirical Approach *Date: *Friday, Sep 29th, 2023 *Time:* 10:00-11:00am *Zoom link:* https://virginiatech.zoom.us/j/89895834017?pwd=ZTFtZXZLMHRwdVhWN1pPZStEdkhT… Meeting ID: 898 9583 4017 Passcode: 090230 *Abstract:* Developers often spend much of their valuable development time on software debugging and bug finding. In addition, software defects cost the software industry as a whole hundreds or even a trillion of US dollars. As a result, many fault localization (FL) techniques for localizing bugs automatically, have been proposed. Despite its popularity, adopting FL in industrial environments has been impractical due to its undesirable accuracy and high runtime overhead cost. Motivated by the real-world challenges of FL applicability, this dissertation addresses these issues by proposing two main enhancements to the existing FL. First, it explores different strategies to combine a variety of program execution information with Information Retrieval-based fault localization (IRFL) techniques to increase FL’s accuracy. Second, this dissertation research invents and experiments with the unconventional techniques of Instant Fault Localization (IFL) using the innovative concept of triggering modes. Our empirical evaluations of the proposed approaches on various types of bugs in a real software development environment shows that both FL’s accuracy is increased and runtime is reduced significantly. We find that execution information helps increase IRFL’s Top-10 by 17–33% at the class level, and 62–100% at the method level. Another finding is that IFL achieves as much as 100% runtime cost reduction while gaining comparable or better accuracy. For example, on single-location bugs, IFL scores 73% MAP, compared with 56% of the conventional approach. For multi-location bugs, IFL’s Top-1 performance on real bugs is 22%, just right below 24% that of the existing FL approaches. We hope the results and findings from this dissertation help make the adaptation of FL in the real-world industry more practical and prevalent. Cheers, Na -- *Na Meng, Associate Professor* Virginia Tech | Computer Science (0902) Gilbert Place, Suite 4311A 220 Gilbert Street, Blacksburg, VA 24060 (540) 231-5492 | nm8247(a)vt.edu

1 0

Ying Zhang's PhD dissertation defense scheduled at 12:30pm on 9/20/23
by Na Meng 18 Sep '23

18 Sep '23

Hi all, You are warmly welcomed to attend Ying's PhD dissertation defense. *Title:* Automatic Detection, Repair, and Demonstration of Vulnerabilities Related to Java API Insecure Usage *Location:* 220 Gilbert Street, Room 4307 *Date:* Wednesday, Sep 20th *Time:* 12:30pm-1:30pm *Zoom link:* https://virginiatech.zoom.us/j/7102599932 Meeting ID: 710 259 9932 Passcode: eGN4c1ErTDMrNlVMQm92MXE0R0xtUT09 *Abstract:* The Java platform and third-party open-source libraries provide various APIs to facilitate secure coding. However, using these APIs securely can be challenging, especially for developers who aren't trained in cybersecurity. Prior work shows that many developers use APIs insecurely, consequently introducing vulnerabilities in their software. Despite the availability of various tools designed to identify API insecure usage, it is still unclear how well they help developers with secure coding practices. This dissertation focuses on (1) exploring the strengths and weaknesses of the existing automated detection tools for API-related vulnerabilities, and (2) creating better tools that detect, repair, and demonstrate these vulnerabilities. We first systematically evaluated the strengths and weaknesses of the existing automated API-related vulnerability detection tools. We observed that most existing tools merely report misuses, without suggesting any customized fixes. Additionally, developers often reject tool-generated vulnerability reports due to their concerns about the correctness of detection, and whether the reported vulnerabilities are truly exploitable. To address the limitations found in our study, the second work proposed a novel example-based approach SEADER, to detect and repair API insecure usage. The third work leverages ChatGPT-4.0 to automatically generate security test cases, and to demonstrate how vulnerable API usage facilitates the supply chain attacks to given software applications. Cheers, Na -- *Na Meng, Associate Professor* Virginia Tech | Computer Science (0902) Gilbert Place, Suite 4311A 220 Gilbert Street, Blacksburg, VA 24060 (540) 231-5492 | nm8247(a)vt.edu

1 0

Research Participants Wanted: Collaborative Data Science Study
by Yunis Hussein 15 Sep '23

15 Sep '23

Hello, We are looking for participants in a user study about collaboration in data science endeavors. As part of this study, you will complete a pre-study survey, a collaborative data analysis task, a post-task survey, and potentially participate in an interview session. During the pre-study survey, you will answer questions about your prior experience with Python, computational notebooks, and collaborative data science. If you qualify, based on the pre-study survey, you will then be placed into a group and asked to submit your availability for 4 one hour sessions to complete a data analysis task. You will then be assigned times to complete the collaborative data analysis task based on your stated availability. After you complete the task, you will fill out a survey regarding your experiences with the task, as well as the software and space used for the task. The task and post-task survey combined is expected to take no more than 4 hours and 30 minutes over the course of about 4 one-hour sessions. You will be compensated $40 in the form of an Amazon gift card. To participate, you must have prior experience with Jupyter Notebooks and data analysis in Python. Participation in this study is voluntary, and any data you provide will be kept confidential. You must complete both the task and survey to receive the compensation. If you are interested, please complete this pre-study survey to determine if you are eligible. Additionally, if you have any questions, please contact Yunis Hussein at yhussein26(a)vt.edu. Link to pre-study survey: https://virginiatech.questionpro.com/t/AXe9DZzU14 Thank you for your time, Yunis Hussein

1 0