
Keval Morabia
github.com/kevalmorabia97 • linkedin.com/in/kevalmorabia97
EDUCATION
UNIVERSITY OF ILLINOIS AT URBANA-CHAMPAIGN (UIUC) Urbana, IL
Master of Computer Science (GPA: 4/4) Aug 2019 - Dec 2020
Coursework: Computer Vision, Machine Learning in NLP, Advanced Information Retrieval
BIRLA INSTITUTE OF TECHNOLOGY AND SCIENCE (BITS), PILANI Hyderabad, India
B.E. (Hons.) in Computer Science (GPA: 9.72/10) Aug 2015 - May 2019
Electives: Artificial Intelligence, Machine Learning, Information Retrieval, Data Mining, Number Theory, Cryptography
RESEARCH EXPERIENCE
UIUC - FORWARD DATA LAB Urbana, IL
Research Assistant Aug 2019 - Mar 2020
• Worked with Prof. Kevin Chang on a novel Visual Attention based Sequence Model in PyTorch for Webpage Information
Extraction (IE) by exploiting contextual information
• Created the largest public labeled dataset of 7.7k webpage screenshots for Product IE
• Created Attention Visualizations for interpretability of results
• Achieved an Accuracy of 94.5% for product Price extraction (Improvement of 16%). Submitted to ECCV 2020
MICROSOFT RESEARCH Bengaluru, India
Research Intern Jan 2019 - Jul 2019
• Implemented Graph Recurrent Neural Networks from scratch in TensorFlow to Learn Embeddings for 300,000 entities in a
heterogeneous graph
• Collaborated with a team of 10 members to design a novel Deep Neural Net architecture
• Utilized embeddings for recommending important messages to a user posted in a channel on MSTeams application
BITS PILANI Hyderabad, India
Research Assistant Jan 2018 - Dec 2018
• Analyzed Twitter Stream for Event Detection by leveraging Wikipedia information
• Segmented tweets and hash-tags; applied Jarvis-Patrick clustering; and summarized newsworthy events
• Achieved a precision of 88.1% (Improvement of 8%)
PROFESSIONAL EXPERIENCE
ARCESIUM INDIA PRIVATE LIMITED Hyderabad, India
Software Engineering Intern May 2018 - Jul 2018
• Enhanced Expense Management System by adding back-end services in Java to modify in MS SQL Server database
• Automated comparison of budgeted inputs with actual expenses on a UI implemented in JavaScript
• Wrote about 100 unit test cases in JUnit and Mockito and increased test coverage by 5%
• Interacted with clients to cater to specific feature requirements
PUBLICATIONS
• Morabia, K., Bhanu Murthy, N. L., Malapati, A., & Samant, S. (2019, June). SEDTWik: Segmentation-based Event Detection
from Tweets using Wikipedia. In Proceedings of the 2019 Conference of the North American Chapter of the Association for
Computational Linguistics (NAACL): Student Research Workshop (pp. 77-85) [Paper] [Code]
INDEPENDENT PROJECTS
• Created pyTweetCleaner, a Python module for cleaning and parsing of Twitter JSON data
• Built an application in JavaFX for Frequent Patterns & Association Rule Mining using Apriori algorithm
SKILLS
• Programming: Java, Python, C, SQL, HTML, CSS, JavaScript, LaTeX
• Technical: PyTorch, TensorFlow, GIT, Java Swing, JavaFX, JUnit, Mockito, Spring Framework, My Batis
LEADERSHIP AND HONORS
• Awarded Merit Scholarship (Top 1% out of 700 undergrad students) Aug 2015 - May 2019
• Organized cultural dance events as President of 150-membered Gujarati Association - BITS Pilani,
Hyderabad; attended by over 1000 students; raised INR 30,000 fund
Jul 2017 - May 2018
评论0