Hi, I'm

Divij Bhatia.

I'm an AI-Enthusiast and a Full-Stack Web Developer who loves to bring ideas to life.

Download Resume
Intro Vector
background
Divij Bhatia

Hi 👋🏻

I'm a master's student at the University of Southern California. Currently, I work at the USC Information Sciences Institute under the supervision of Dr Pedro Szekely.

My interest lies in solving complex problems that leverage technology and available resources.

Talking about my experience with different technologies, I have worked with Java Server Pages, Servlets, J2EE Enterprise Java Beans and Flask framework to build scalable web applications. I also have some experience of using Artificial Intelligence to solve some common problems like text summarization, scene understanding, question answering etc.

Being an avid learner, I love exploring new technologies which help me to build better software with clean, efficient and scalable code. Adding to that, I am a firm believer of experiential learning and I'm always open to working with people who bring out the best in me.

Apart from work, I love to watch TV shows and to photograph this world on my camera.

education
Jan 2019 - Present

University of Southern California, Los Angeles, CA

USC Viterbi School of Engineering Master of Science, Computer Science GPA: 3.50/4.00
July 2014 - June 2018

Jaypee Institute of Information Technology, Noida, India

Bachelor of Technology, Computer Science & Engineering CGPA: 8.9/10.0
experience
May 2019 - Present

USC Information Sciences Institute

Student Worker Roles & Duties:

Developed T2WML - a cell-based language to map tabular data to Wikidata records

Collaborated with a frontend developer to build a web application based on Flask server for users to leverage T2WML

Analyzed and optimized runtime processing of T2WML; reduced data processing delay by 10x

May 2017 - July 2017

Niki.ai

Technology Intern Roles & Duties:

Integrated Unified Payments Interface (UPI) as a new and secured pay-through-chat service

Implemented a recommender system to suggest shuttle boarding points to users; reduced time to book a shuttle by 20%

Increased expressiveness of Niki chatbot by adding support for UTF-8 emoticons; improved user experience and engagement by 10%

skillset
Here are the domains I have been working on from past few years.
Here are the skills that I have acquired in the above mentioned domains.
projects
May 2019 - Present

T2WML- Table to Wikidata Markup Language

Project pursued at USC Information Sciences Institute

It's a cell-based language to map tabular data to Wikidata data model. It is designed to identify and map data and their context qualifiers in arbitrary data layouts found in Excel and CSV files without the need of complex preprocessing steps to transform tables into a canonical "Database" representation.

  • Compiler Design
  • Knowledge Graphs
  • Wikidata
  • Python 3
  • Lark Parser
  • Flask
Aug 2017 - May 2018

AI Game Agent

An AI-based game agent based on minimax adversarial search with alpha-beta pruning algorithm for a two-player zero-sum game.

  • Artificial Intelligence
  • Adversarial Search
  • Python 3
Aug 2017 - May 2018

Scene Understanding for Visually Impaired People using Deep Learning

BTech Major Project

I built this application for the visually impaired people to help them observe the scene around them. The app uses deep learning and computer vision algorithms to analyze the visual features of an image or video and generate a vocal response to describe the scene. The tool can detect places (Ex. Classroom, Gymnasium, Park etc.), objects and people and is also capable of predicting their gender and age.

  • Deep Learning
  • Computer Vision
  • Supervised Learning
  • Convolutional Neural Networks
  • Python 3
  • Keras
  • PyTTS
Jan 2017 - May 2017

Extractive Text Summarization using Word Vector Embedding

BTech Minor Project - II

This application focuses on providing a solution to the extractive text summarization problem using word vector embedding and machine learning algorithms. The algorithm has Rouge-1, Rouge-2 and Rouge-L F-Scores equal to 0.38249, 0.2256 and 0.27486 respectively (tests performed on DUC 2002 Dataset).

  • Machine Learning
  • Natural Language Processing
  • Supervised Learning
  • Multilayer Perceptron
  • Python
  • SciPy
  • NLTK
  • Flask
Jan 2017 - May 2017

Remarque ― Cloud Based Notes Taking Application

A simple Java-based SaaS application for taking notes. The application uses concepts of J2EE and uses Enterprise Java Beans for processing data requests.

  • Cloud Computing
  • J2EE
  • EJB
  • JavaServer Pages
  • Servlets
  • AJAX
  • HTML
  • CSS
  • JavaScript
Aug 2016 - Dec 2016

Quewer ― Factoid Question Answering System

BTech Minor Project - I

A web-based application for answering questions whose answer is a concise fact. The application uses ML & NLP algorithms to analyze and understand the question and uses DBpedia knowledge base to answer it.

  • Machine Learning
  • Natural Language Processing
  • Supervised Learning
  • Knowledge Representation
  • Semantic Web
  • Support Vector Machines
  • DBpedia
  • SPARQL
  • Java
  • Java Swing
Jan 2016 - May 2016

Mobile Phone Recommender System

A web application which suggests different mobile handsets to users based on the tasks for which they use their mobile phones. Recommendations were made using a scoring system which gave a cumulative score to every feature of a phone based on the user reviews.

  • Natural Language Processing
  • Machine Learning
  • Unsupervised Learning
  • K-Means Algorithm
  • Web Crawling & Scraping
  • Jsoup
  • Java
  • JSP
  • AJAX
  • HTML
  • CSS
  • JavaScript
Dec 2015 - Jan 2016

ebook-shala ― Online Book Store

This online library provides a plethora of books, divided into different categories based on stream, course etc., to download. The application also features a local search engine to search a book using its name, author name, ISBN number, stream etc. I developed this website to learn how a web application works.
  • HTML
  • CSS
  • JavaScript
  • JavaServer Pages
  • AJAX
  • MySQL
  • JDBC
research
19-21 Nov 2019

T2WML: Table to Wikidata Mapping Language

Tenth International Conference on Knowledge Capture (K-Cap 2019) Marina del Rey, California, United States Pedro Szekely, Daniel Garijo, Divij Bhatia Jiasheng Wu, Yixiang Yao and Jay Pujara
26-30 Oct 2019

T2WML: A Cell-Based Language To Map Tables Into Wikidata Records

Eighteenth International Semantic Web Conference (ISWC 2019) The University of Auckland, New Zealand Pedro Szekely, Daniel Garijo, Jay Pujara, Divij Bhatia and Jiasheng Wu
14-15 Dec 2017

Extractive Text Summarization using Word Vector Embedding

International Conference on Machine Learning & Data Science (IEEE ICMLDS 2017) Bennett University, India Aditya Jain, Divij Bhatia, Manish K Thakur View Paper
volunteer experience
Aug 2017 - May 2018

Developer Student Club JIIT Noida

Advisor Roles & Duties:

Formed a new team of coordinators by interviewing candidates

Mentored and counseled coordinators for the smooth functioning of the organization

July 2016 - Aug 2017

Developer Student Club JIIT Noida

Coordinator Roles & Duties:

Started a new DSC chapter at sector-128 campus of JIIT

Interviewed candidates for the sector-128 DSC team

Conducted workshops on Web Development and Android App Development

Organized annual technical fest - DevFest 2016, hackathons, tech talks etc.

Mentored sophomores on their web-development project

July 2016 - Aug 2017

Knuth Programming Hub JIIT Noida

Coordinator Roles & Duties:

Conducted workshops on basic algorithms for freshmen & sophomores

Organized solo and team coding competitions

Mentored freshmen to get started with programming

testimonials
blog