Concert Demand Forecasting System
Concert ticket demand forecasting system using machine learning and large language models (LLM). Analyzes social media, news, historical data, and trends for accurate sales prediction.
Enter artist details and click "Predict Demand"
Data collection from multiple sources: Spotify, YouTube, social media, news sites, ticket platforms.
LLM integration (OpenAI, YandexGPT, Google Gemini) for sentiment analysis and event extraction from news.
Machine learning models (CatBoost, XGBoost, LightGBM) for demand forecasting and sellout probability.
REST API for predictions, artist analysis and real-time report generation.
Interactive dashboard for visualizing predictions, top artists and historical analysis.
Task scheduler for automatic data updates, model retraining and monitoring.
Comment sentiment analysis using LLM. Identifies categories: excitement, criticism, question, neutral. Evaluates purchase intent.
{
"sentiment_score": 8.5,
"intent_to_buy": true,
"category": "excitement",
"buy_intent_percentage": 67%
}
Extracting significant events from news: album releases, tour announcements, scandals, awards. Impact assessment on popularity.
{
"event_type": "album_release",
"impact_score": +8,
"confidence": 0.95
}
Demand prediction based on 200+ features. Uses ensemble of CatBoost + XGBoost models for maximum accuracy.
{
"demand_score": 85,
"sellout_probability": 92%,
"estimated_time": "<24 hours"
}
Automatic generation of analytical reports with recommendations for promoters and secondary market forecasts.
1. Executive Summary 2. Strengths 3. Risks and Limitations 4. Recommendations 5. Secondary Market Forecast
/predict
Concert demand prediction
/artist/{name}
Detailed artist analysis
/top-artists
Top artists by demand
/health
System status check