Sentiment-Analysis: A Jupyter Notebook repository from ajaykumar8

# Sentiment Analysis

## Overview

This project implements a sentiment analysis tool that processes text data, specifically movie reviews and tweets, to determine their sentiment scores. The tool utilizes a lexicon-based approach, leveraging a predefined sentiment lexicon to compute sentiment scores for input text.

## Features

- **Data Loading**: Loads IMDB movie reviews and live Twitter data for analysis.
- **Tokenization**: Tokenizes text by removing punctuation and stop words.
- **Sentiment Scoring**: Calculates sentiment scores based on a sentiment lexicon.
- **Data Visualization**: Analyzes sentiment across different tweets to summarize public opinion.

## Requirements

- Python 3.x
- Libraries:
  - `nltk`
  - `json`
  - `urllib`
  - `tkinter`

You can install the necessary libraries using pip:

```bash
pip install nltk

Usage

Clone the repository:

git clone https://github.com/ajaykumar8/sentiment-analysis.git
cd sentiment-analysis

Download the necessary NLTK data files:

import nltk
nltk.download('stopwords')
nltk.download('punkt')

Prepare the lexicon file (lexicon.txt) with sentiment scores. Ensure the format is correct as follows:
```
word1   score1
word2   score2
...
```

Prepare a JSON file containing tweets in the following format:

[
  {"text": "Tweet text 1"},
  {"text": "Tweet text 2"},
  ...
]

Run the script to perform sentiment analysis:
```
python sentiment_analysis.py
```

Functions

`tokenize(doc)`

Input: A string of text (document).
Output: A list of tokens (words) after removing punctuation and stop words.

`sentiment_score(doc)`

Input: A string of text (document).
Output: A dictionary containing words and their corresponding sentiment scores from the lexicon.

`score_message(doc)`

Input: A string of text (message).
Output: The overall sentiment score for the message.

Contributing

Contributions are welcome! If you have suggestions for improvements or bug fixes, please create a pull request or open an issue.

License

This project is licensed under the MIT License - see the LICENSE file for details.

ajaykumar8/Sentiment-Analysis