Natural Language Processing - CSE-538 - Course Project
Contributors: Anurag Dutt and Srinidhi Bhat
Aim: Main aim of this project is to determine the movement of the stocks of the top 500 S&P companies listed in the US Stock Market by using Analyst reports called 8-k filings.
The project ensures to predict the sentiment using no "numerical metric" and only sentences.
Techniques used:
- A Bi-LSTM module was developed and used for dependency parsing. The features extracted was fed to seperate CNN,RNN,MLP and CNN-RNN modules.
- In our experiment we found CNN-RNN to perform the best.