Russian Sentiment Analysis Bot
Year: 2025
Technologies: Python, Flask, Hugging Face Transformers, Bootstrap 5, JavaScript, NewsAPI
Overview
The Russian Sentiment Analysis Bot is a full-stack Flask web application that analyses sentiment in Russian-language text using artificial intelligence. It combines user-inputted text analysis with live Russian news sentiment scoring, using Hugging Face’s RuBERT model (sismetanin/rubert-ru-sentiment-rusentiment).
Key Features
- AI-powered sentiment classification for positive, neutral, or negative text.
- Real-time Russian news headline analysis using NewsAPI.
- Russian flag–themed Bootstrap interface with live loading indicators and colour-coded results.
- Instant feedback with asynchronous JavaScript updates.
- Deployable full-stack Flask and Bootstrap application, easily adaptable for other languages or data sources.
Technical Details
The backend is written in Python (Flask), with a RESTful API that connects to the RuBERT sentiment model hosted on Hugging Face for text inference. The frontend uses Bootstrap 5 and JavaScript for an interactive experience, with colour-coded feedback and live news integration via NewsAPI. Results are returned as structured JSON with polarity and confidence scores.
Rationale & Purpose
This project demonstrates how multilingual NLP and machine learning can be applied in open-source intelligence to assess sentiment in foreign-language media environments. It was developed as part of an exploration into Russian-language data analysis and media monitoring relevant to UK intelligence and defence applications.