Final Project for CSE-8803 Deep Learning for Text Data, Fall '21. This project was completed along with Micaela Siraj and Praveen Balaji
The TASTY paper seeks to prevent doomscrolling through restaurant reviews by generating review summaries with the most relevant information. Doom- scrolling refers to the seemingly endless page scrolling while on a laptop or mobile device. Team 15 worked to develop an abstractive, classification summarizer that generates review summaries by both star and topic category. The goal is for the summaries to contain the most pertinent information from the overall group of reviews.