Decision Critical Bayesian Networks

Application for the identification of decision-critical situations through dynamic bayesian networks.

This project is in its core a Progressive Web App used for modeling dynamic bayesian networks. With the backend serving inference calculations through ActiveMQ for those modeled networks.
Currently, DCBN is designed with the observation of maritime space in mind and is used to rapidly and in realtime evaluate situations involving vessels. The main features are:

• User System with two privilege groups
• Network editor to model dynamic bayesian networks
• Network playground to experiment with the networks
• Evidence editor to define expressions converting metadata to evidence used by the networks
• Backend integration with ActiveMQ to calculate inference for given metadata

Repo Structure

Docs

The documentation folder contains:

• The project specifications
• Any other non-code-related files

Frontend

The frontend is a PWA build with Vue, TypeScript, Vuetify, and visNetwork.
Using npm for dependency management.

Backend

The backend is for most parts a REST-Service build using Spring Boot and AMiDST for inference.
Using maven for dependency management.