Stock Price Prediction Using Twitter Sentiment

This project aims to predict the stock price of Reliance Industries Limited (RIL) using Twitter sentiment analysis. By combining social media signals with historical stock prices, we attempt to explore the impact of public sentiment on market behavior.

Overview

The project pipeline includes:

• Collecting tweets relevant to Reliance
• Performing sentiment analysis on tweets
• Merging sentiment scores with historical RIL stock data
• Building a predictive model (Linear Regression)
• Visualizing the effect of sentiment on stock prices

Datasets Used

1. Tweets.csv: Contains raw tweets about Reliance.
2. RIL.csv: Historical stock price data of Reliance Industries.
3. Twitter_Dataset.pkl: Pickled version of the preprocessed tweet dataset.

Methodology

1. Preprocessing Tweets

• Clean tweet texts by removing mentions, URLs, emojis, and special characters
• Convert to lowercase and remove stopwords
• Perform tokenization and lemmatization

2. Sentiment Analysis

• Each tweet is assigned a polarity score using TextBlob
• Tweets are labeled as Positive, Neutral, or Negative

3. Feature Engineering

• Daily sentiment scores are averaged
• Resulting features include daily average sentiment and counts of each sentiment class

4. Data Merging

• Stock price data is merged with daily sentiment scores
• Final dataset includes stock closing prices and aggregated sentiment data

5. Predictive Modeling

• A linear regression model is trained to predict next-day closing prices
• Features used: sentiment polarity, sentiment counts

6. Evaluation

• Performance is evaluated using Mean Squared Error (MSE)
• Visual plots compare predicted vs actual stock prices

Results

The model shows that Twitter sentiment holds weak but noticeable predictive power for RIL stock movements. The results validate that sentiment data can enhance traditional price forecasting.

Dependencies

• pandas
• numpy
• matplotlib
• seaborn
• textblob
• sklearn

Install them using:

pip install -r requirements.txt

How to Run

1. Clone the repository or download the .ipynb file
2. Ensure all datasets (RIL.csv, Tweets.csv) are in the same directory
3. Open the notebook and run all cells in order

Credits

This project is developed as part of a stock prediction research case study using NLP and sentiment-driven market signals.