Traffic Event Detection using Twitter Data

This repository contains the the implementation of machine learning based traffic inident detection using multi-modal data (Twitter data + Sensor data).

System architecture

Available modules and scripts

• main.py - Train and run the event detector
• tweets_fetch.py - Fetches tweets from twitter using twitter api
• filter_tweets.py - tweets preprocessing
• models - Defines our deep learning models
• training_ML - to trian via ML models (decision tree, Random Forest, XGboosting...)
• train_detector - module for training the detector

Installation and setup

Running the traffic-event detector

To run the detector:

python src\main.py