I am a fourth-year PhD Student in Computer Science at Virginia Tech, Blacksburg. I am part of the DLRL group, where Dr. Edward A. Fox advises me. I'm also affiliated with the Center for Human-Computer Interaction and Sanghani Center for Artificial Intelligence & Data Analytics at Virginia Tech. I plan to graduate in Summer 2024.
I am broadly interested in building Deep Learning & NLP models in a low-resource setting. My recent research focuses on developing knowledge graphs for heterogeneous long scholarly documents (ETDs) and imbibing the learnings into the DL/NLP models for better predictions.
On the application side, I am primarily focused on building intelligent systems for understanding & processing textual data originating from clusters of heterogeneous documents (such as theses, dissertations, reports, etc.) in the scholarly domain and providing guidance (Question-Answering, and Recommendations) with explainability to increase trust & reliability.
I am broadly interested in building Deep Learning & NLP models in a low-resource setting. My recent research focuses on developing knowledge graphs for heterogeneous long scholarly documents (ETDs) and imbibing the learnings into the DL/NLP models for better predictions.
On the application side, I am primarily focused on building intelligent systems for understanding & processing textual data originating from clusters of heterogeneous documents (such as theses, dissertations, reports, etc.) in the scholarly domain and providing guidance (Question-Answering, and Recommendations) with explainability to increase trust & reliability.
Recent News
- June 2023: Our paper "Integrated Digital Library System for Long Documents and their Elements" was nominated for the Best Student Paper Award
- March 2023: Long paper accepted at JCDL-2023
- February 2023: Co-inventor on U.S. Provisional Patent Application No. 63/448,159, filed 24 February 2023
- May 2022: Started as an NLP Intern at Deloitte Data Science Group
- June 2021: Filed a US patent application (Co-inventor)
- February 2021: Passed PhD Qualifiers
- November 2020: Presented (virtually) our paper on Aspect Classification for Legal Depositions at JURISIN'20
Publications
Integrated Digital Library System for Long Documents and their Elements
Satvik Chekuri, Prashant Chandrasekar, Bipasha Banerjee, Sung Hee Park, Nila Masrourisaadat, Aman Ahuja, William Ingram, Edward Fox
(To-appear) JCDL, 2023
Quantification of Gender-related Stereotypes in Psychotherapy Sessions
Sagar Mahesh Badve, Satvik Chekuri, Wenqi Shen, Alan Wang
Pre-ICIS SIGDSA, 2022
Aspect Classification for Legal Depositions
Saurabh Chakravarty, Satvik Chekuri, Maanav Mehrotra, Edward A. Fox
International Symposium on Artificial Intelligence, 2020 Oct 14, pp. 179-195
Get the Job! An Immersive Simulation of Sensory Overload
Leonardo Pavanatto, Feiyu Lu, Shakiba Davari, Emily Harris, Anthony Folino, Samat Imamov, Satvik Chekuri, Leslie Blustein, Wallace S. Lages, Doug A. Bowman
2020 IEEE Conference on Virtual Reality and 3D User Interfaces Abstracts and Workshops (VRW), 2020 Feb 21, pp. 509-510
Elasticsearch (ELS) CS5604 Fall 2019
Yuan Li, Satvik Chekuri, Tianrui Hu, Soumya Arvind Kumar, Nicholas Gill
2019 Nov 11
Projects
This tool will aid the annotators in the creation of a Knowledge Base that is rich with topics/keywords and Question-Answers for each chapter in ETDs.
Interactive Text Summarization using Explorable MMR
Maximal Marginal Relevance (MMR) is a rank-based technique for producing summaries. This project aims to explore the MMR technique by incorporating the human-in-the-loop and providing an interactive text summarization system.
Understanding the Impact of Graph Diffusion on Robust Graph Learning. We propose graph diffusion convolution as a defense mechanism against adversarial perturbations in graph learning tasks. Our experiments show up to 8% improvements in accuracy.
Visualization of the impact of COVID-19 coupled with census demographics
Understanding how COVID-19 has impacted different Point of Interests (POIs) by analyzing the user behaviour (such as visits, time spent etc) and mapping it with the demographic information from US Census data.
