HeadlineWriter

Collaborator: Varnith,Naren, Ranvir

Tools Used: Pytorch , Pysark, nltk, BeautifulSoup, gensim

1. Obtained~700,000 news articles from various data sources, by building a web scrapper on Python.
2. Tokenized words and visualized text patterns on Pyspark, giving insights on the content of news present.
3. Built a Sequence to Sequence LSTM, to generate headlines for given articles, optimizing it using Stochastic gradient descent.

Based on:https://nlp.stanford.edu/courses/cs224n/2015/reports/1.pdf