About

I am a Computer Science PhD student in the School of Interactive Computing at Georgia Institute of Technology.

I am a research assistant at the Entertainment Intelligence Lab advised by Dr. Mark Riedl and a teaching assistant for CS 3600 (Introduction to AI). My research involves everything from using natural language processing and deep learning techniques to create an AI agent that can collaboratively compose a story to playing text-adventure games using deep reinforcement learning and knowledge graphs.

Most of my research interests in the field of AI and machine learning lie in natural language processing, computational creativity, and reinforcement learning. I am especially interested in applying reinforcement learning to language.

Bio

I am originally from Visakhapatnam in Andhra Pradesh, India. Since then, I've moved around quite a bit. I've lived in Singapore, Hong Kong, New Jersey, California, and multiple places in India such as Chennai, Hyderabad, and Bangalore.

I was recently with the Reinforcement Learning team at Microsoft Research Redmond and worked on using deep reinforcement learning and knowledge graphs to play text-adventure games. I have worked as a Software Engineering Intern at Oracle Intelligent Bots, where I developed a patent pending algorithm to generate noisy knowledge graphs for low resource natural language datasets and a method to use them to improve the natural language understanding capabilities of the chatbot platform. I have previously worked as a Machine Learning Intern at Radix Health, developing and deploying products such as no-show risk models and chatbots that help clinics manage their schedules more efficiently in addition to improving patient access.

My hobbies include Tae Kwon Do, camping/hiking, reading, and gaming. I am a 1st Dan black belt in Tae Kwon Do. In my leftover spare time, I go out into nature, read sci-fi and high fantasy novels, and play MMORPGs.

Publications

2019

Story Realization: Expanding Plot Events into Sentences
P Ammanabrolu, E Tien, W Cheung, Z Luo, W Ma, LJ Martin, and MO Riedl
Preprint.
Code

Toward Automated Quest Generation in Text- Adventure Games
P Ammanabrolu, W Broniec, A Mueller, J Paul, and MO Riedl
Computational Creativity in Natural Language Generation @ INLG 2019. Tokyo, Japan.

Interactive Fiction Games: A Colossal Adventure
M Hausknecht, P Ammanabrolu, MA Cote, and X Yuan
Deep Reinforcement Learning Workshop @ NeurIPS 2019. Vancouver, BC.
Code

Transfer in Deep Reinforcement Learning using Knowledge Graphs
P Ammanabrolu and MO Riedl
Text Graphs @ EMNLP 2019. Hong Kong.
Code

Guided Neural Language Generation for Automated Storytelling
P Ammanabrolu, E Tien, W Cheung, Z Luo, W Ma, LJ Martin, and MO Riedl
StoryTelling@ACL 2019, Florence, Italy.
Code, Slides

Playing Text-Adventure Games with Graph-Based Deep Reinforcement Learning
P Ammanabrolu and MO Riedl
NAACL-HLT 2019, Minneapolis, MN.
Code, Slides, Blog

2018

Playing Text-Adventure Games with Graph-Based Deep Reinforcement Learning
P Ammanabrolu and MO Riedl
NeurIPS 2018 Wordplay: Reinforcement and Language Learning in Text-based Games Workshop, Montreal, QC.

Event Representations for Automated Story Generation with Deep Neural Nets
LJ Martin, P Ammanabrolu, X Wang, W Hancock, S Singh, B Harrison, and MO Riedl
Thirty-Second AAAI Conference on Artificial Intelligence (AAAI-18), New Orleans, LA.

2017

Improvisational Storytelling Agents
LJ Martin, P Ammanabrolu, X Wang, S Singh, B Harrison, M Dhuliawala, P Tambwekar, A Mehta, R Arora, N Dass, C Purdy, and MO Riedl
NeurIPS 2017 Workshop on Machine Learning for Creativity and Design, Long Beach, CA.

Event Representations for Automated Story Generation with Deep Neural Nets
LJ Martin, P Ammanabrolu, W Hancock, S Singh, B Harrison, and MO Riedl
KDD 2017 Workshop on Machine Learning for Creativity, Halifax, NS.

Projects

  • Health Desk at HackIllinois - Desktop app to detect posture using computer vision; built with OpenCV and wxPython
  • Labyrinth at SwampHacks - 3D survival maze game built with Unity engine and C#
  • Other video game projects such as Time before Time from working with VG Dev Club at Georgia Tech and a Rougelike written in C++ from earlier.