TY  - BOOK
AU  - Sanghani, Sureshkumar 
AU  - Brahma, Dweepobotee 
TI  - A Twitter sentiment based Indian stock price forecasting using deep learning
U1  - 006.312 
PY  - 2023///
CY  - IIT Jodhpur
PB  - Department of Computer Science and Technology
KW  - Department of Computer Science and Technology
KW  - Stock Index Prediction
KW  - Sentiment Analysis
KW  - MTech Theses
N1  - A share market investor is always seeking advice from different investment experts to decide their next investment plan or strategy. Basically, they always look for news about a company’s (in which they are interested) future plans, current financial position, current and future order book, and the impact of local or global scenarios and local or international government policies, etc. They must familiarize themselves with all this information as well as the different opinions of various experts. Twitter is a very good source of such news or expert advice on this topic. Almost all major financial news agencies and stock market experts have their Twitter handles, which publish news and their opinions on the share market based on recent events.

We will scrape these tweets and aggregate news about Nifty Bank, Nifty IT, Nifty Energy, and Nifty Automobile stock indexes and do sentiment analysis on this data to calculate the sentiment of each day. We will use stock data of all 4 indexes and use this sentiment value of a given day as one additional variable to train a deep learning model that can predict day-level stock prices at the end of the day. The stock markets are often volatile and change abruptly due to economic conditions, political situations, and major events. Therefore, including the effect of some major events for different top stock indexes is worthwhile for the model to learn the impact and predict the price more accurately
ER  -