Siva Uday Sampreeth Chebolu
Dedicated AI/ML Researcher and a Ph.D. graduate with a solid background in computer science, focused on scalable machine learning and efficient data solutions. Prior industry experience at Amazon and HPE, with a track record of scholarly contributions through multiple publications. Contributed to improvements in software performance and developed innovative approaches within collaborative team settings in both tech industry internships and academic research. Eager to leverage this blend of academic and practical expertise to excel in new Research Scientist opportunities.
ABOUT ME
I just completed my Ph.D. in Computer Science at the University of Houston. I worked under Dr. Thamar Solorio at the Research in Text Understanding and Analysis of Language (RiTUAL) lab. My research interests lie in the field of Natural Language Processing, with a special focus on aspect-based sentiment analysis and multimodal aspect-based sentiment analysis.
EDUCATION
University of Houston
Ph.D. in Computer Science, Aug 2019 - Dec 2023
Advisor: Dr. Thamar Solorio
University of Houston
M.S. in Computer Science, Aug 2017 - May 2019
Advisor: Dr. Carlos Ordonez
Gayatri Vidya Parishad College of Engineering
B.Tech in Computer Science, Oct 2013 - Apr 2017
RESEARCH
Research Interests: Areas that deal with Natural Language Processing, Machine Learning, Data Science, and Deep Learning excite me the most and bring out the best in me
Aspect Based Sentiment Analysis
Followed a simple multi-task strategy to jointly solve all the subtasks of aspect-based sentiment analysis
Implemented various models including LSTMs, GRUs, and CNNs deep neural networks on restaurant domain datasets from SemEval competitions, urban neighborhoods domain from SentiHood, and several other datasets
Leveraged several Transformer architectures including BERT, T5, and RoBERTa to jointly detect Aspect Categories, Aspect Terms, and their polarities in Aspect-Based Sentiment Analysis
Familiarized with distributed training on GPUs/TPUs for running deep learning models
Aspect Category Detection
Conducted research on detecting aspect categories from review texts of several domains
Improved the performance of detecting aspect categories by leveraging a multi-task approach with aspect terms
Did a survey of all the methods to detect the aspect categories and summarized the observations, methods, limitations, and pointed out the future directions
Scalable Machine Learning in the R Language using Summarization Matrix
Performed Data Segregation on large datasets by dividing into chunks of pre-defined size in R
Implemented a Summarization Matrix, Gamma (Γ), in the Cpp Language capturing fundamental Statistical properties
Computed Machine Learning Models on input data by applying properties of Γ matrix on the obtained chunks
Integrated into a package in the R language - takes a dataset as input and outputs Machine Learning Models and predictions
Eliminated the Physical Memory (RAM) limitations, making it less resource hungry
Enhanced model computation speedup by six times when compared to current best packages in R
Achieved an accuracy greater than 99.9% when compared to the results from other packages in R
EXPERIENCE
Data Science NLP Intern (Summer 2023) - Hewlett Packard Enterprise
Houston, TX Developed an NLP-based Question-Answering system to facilitate seamless user-data interactions.
Leveraged Generative AI and various Large Language Models (LLMs) to create a system providing insightful, narrative responses to a broad array of marketing metrics queries.
Presented the end-to-end solution to senior executives, earning positive feedback for the design and functionality.
Data Scientist Intern (Summer 2022) - Hewlett Packard Enterprise
San Jose, CA Built a scalable and holistic data science and analytics platform for digital ad campaign optimization
Performed exploratory analysis on datasets with > 2.5B records to provide valuable insights for marketing strategies
Used AWS EMR service to facilitate the analysis of data and run ML models on it to optimize the ad campaigns
Achieved 4x speed up of the application and improved the performance of ML models by 2 times
Applied Scientist Intern (Summer 2022) - Amazon
Houston, TXCreated an end-to-end system to improve Amazon.com search by boosting the precision of descriptive queries
Cleaned and transformed huge amounts of customer data using PySpark and SQL
Utilized various deep learning models including LSTMs, GRUs, and Transformer-based models to get the contextual representations of Amazon customer reviews and product descriptions
Did data annotation for a sample of Amazon customer reviews dataset and a product dataset
Junior Application Developer (Sep 2017 - Jun 2019) - Enterprise Systems
Houston, TXPlayed a lead role in the design and development of a Real-Time Messaging System for the University of Houston
Built Restful web services for PeopleSoft integration by deploying a Node.js module on the Twilio Cloud Servers
Achieved user communication via PeopleSoft Integration Broker
Associate Engineer (Mar 2017 - Aug 2019) - Kony India
Hyderabad, IndiaAccomplished extensibility for retail banking applications
Implemented three degrees of freedom [namely App-Id & App-Version, Channel & Device, and Role-Id & User-Id] in an otherwise flat Software Development Kit (SDK) to improve application extensibility
Reduced client-side code dependency to a bare minimum and paved the way for a more server-centric application update
Facilitated application updates to be customizable with the client, given the freedom to define their user interface for the existing application alongside the default customizations
Scaled down the server-client connection requests by 80% by implementing a better version control, Delta Optimization, in the SDLC
PUBLICATIONS
Chebolu, Siva Uday Sampreeth, Franck Dernoncourt, Nedim Lipka, and Thamar Solorio. OATS: Opinion Aspect Target Sentiment Quadruple Extraction Dataset for Aspect-Based Sentiment Analysis, Accepted at COLING/LREC 2024, arXiv:2309.13297
Chebolu, Siva Uday Sampreeth, Franck Dernoncourt, Nedim Lipka, and Thamar Solorio. A Review of Datasets for Aspect-based Sentiment Analysis, 2022, IJCNLP-AACL 2023
Siva Uday Sampreeth Chebolu, Paolo Rosso, Sudipta Kar, and Thamar Solorio. 2022. Survey on Aspect Category Detection. ACM Comput. Surv(May 2022).
Chebolu, Siva Uday Sampreeth, Franck Dernoncourt, Nedim Lipka, and Thamar Solorio. “Exploring Conditional Text Generation for Aspect-Based Sentiment Analysis.”, PACLIC 35, 2021
S. T. Al-Amin, S. Uday Sampreeth Chebolu and C. Ordonez, "Extending the R Language with a Scalable Matrix Summarization Operator," 2020 IEEE International Conference on Big Data (Big Data), 2020, pp. 399-405, DOI: 10.1109/BigData50022.2020.9378399.
Chebolu S.U.S., Ordonez C., Al-Amin S.T. (2019) Scalable Machine Learning in the R Language Using a Summarization Matrix. In: Hartmann S., Küng J., Chakravarthy S., Anderst-Kotsis G., Tjoa A., Khalil I. (eds) Database and Expert Systems Applications. DEXA 2019. Lecture Notes in Computer Science, vol 11707. Springer, Cham. https://doi.org/10.1007/978-3-030-27618-8_19
Chebolu, Siva Uday Sampreeth Chebolu (2019) A General Summarization Matrix for Scalable Machine Learning Model Computation in the R Language (Order No. 28181050). Available from Dissertations & Theses @ University of Houston; ProQuest Dissertations & Theses Global. (2448654831).
RESUME
Sampreeth_Chebolu_Resume.pdf