Persian Sentiment Analysis

Python, Transformers, PyTorch, Flask

Overview

A system for analyzing Persian text and classifying it as positive, negative, or neutral. Designed for social media, website feedback, and customer review analysis.

Features

• Sentiment classification for Persian text
• Fast and accurate text processing
• Can be integrated into web apps or chatbots via API

Technical Details

• Transformer-based model (BERT Persian) for text classification
• Preprocessing for Persian language
• Model accuracy: 91% on a custom Persian text dataset (10k samples)
• Flask API for serving predictions