Sentiment Analysis for Financial News And Tweets

Overview

This project aims to leverage sentiment analysis techniques to analyze financial news articles and social media posts for their impact on stock price movements. By employing Natural Language Processing (NLP) and machine learning algorithms, we aim to determine if sentiment scores from textual data can be used as a predictive signal for trading decisions. This research has the potential to provide valuable insights for investors and traders.

Technologies Used

• Python: Data analysis and visualization are performed using Python programming language.
• Pandas: For data manipulation and analysis.
• sqlite3: For creating database
• Matplotlib and Seaborn: Used for creating visualizations to better understand the trends.
• Kaggle API: Utilized for accessing the dataset.

Kaggle Authentication

To access the dataset from Kaggle, follow these steps:

1. Go to Kaggle Account Settings.
2. Download your Kaggle API key as a kaggle.json file.
3. Place the kaggle.json file inside the /project/ directory.

Filepath: /project/kaggle.json

{
  "username":"a*****h",
  "key":"3a2b***********************25"
}

Set Execute Permissions for the Pipeline Script

Before running the analysis pipeline, ensure that the pipeline script has execute permissions. Run the following command:

chmod +x ./project/pipeline.sh

Run the Analysis Pipeline

Navigate to the project directory and execute the pipeline script:

cd project && ./pipeline.sh

Set Execute Permissions for the Test Pipeline Script

If you want to run the test pipeline, grant execute permissions to the test script:

chmod +x ./project/tests.sh

Run the Test Pipeline

Navigate to the project directory and execute the test script:

cd project && ./tests.sh

Analysis Report

Explore the detailed analysis report here.