Free-Text Summarization and Sentiment Analysis Web App

This project is a Flask-based web application that processes free-text feedback and performs the following tasks:

• Keyword Extraction: Extracts the top 10 keywords along with their frequencies.
• Sentiment Analysis: Uses XLM-Roberta to analyze the sentiment of the feedback (Positive/Negative/Neutral).
• Text Summarization: Summarizes the input text using mBART (for multilingual support) or BART (for English).

Features

1. Language Detection: Automatically detects if the input text is in English or Japanese.
2. Keyword Extraction: Extracts the most relevant keywords based on their frequency.
3. Sentiment Analysis: Provides sentiment classification for the input text.
4. Summarization: Summarizes the input text using advanced NLP models.
5. Multilingual Support: Uses mBART for multilingual summarization (currently supports English and Japanese).

Demo

• Input your feedback in English or Japanese.
• Get the top keywords, sentiment analysis, and summarized version of your text.

Installation

Prerequisites

• Python 3.8+
• pip package manager

Clone the repository

git clone https://github.com/mingleeeeee/free-text-summarization
cd free-text-summarization
python -m venv venv
source venv/bin/activate  # On macOS/Linux
# or
venv\Scripts\activate  # On Windows
pip install -r requirements.txt
python -m spacy download en_core_web_sm  # English model
python -m spacy download ja_core_news_sm  # Japanese model