Technical Description
The invention provides a method and system to determine the cause of traffic congestion in real time using automated text classification of social media posts. The system filters raw social media data and uses two sequential classifiers to identify road traffic relevance and specific non-recurrent causes like Accident, External, Road Works.
Problems Addressed
- Low Coverage of Physical Location Sensors
- High Computational Costs of Infrastructure
- Unpredictability of Non-Recurrent Congestion
- Excessive Irrelevant Data in Social Platforms
- High Computational Load for Real-Time Filtering
Tech Features
- Sensorless Traffic Monitoring
- Lexicon Keyword Pre-Filtering
- Two-Stage Post Classification
- Three-Class Cause Categorisation
- Persistence Time Prediction
- News-Article Cross-Validation
Target Audience
- Smart City Infrastructure Planners
- Traffic Management Authorities
- Navigation Application Providers
Tech ID: P11-2160 TRL 5 Patent Status: Granted Available For Exclusive and Non-exclusive License
