This repository provides a logging component of the multi-level explainability framework for multi-agent BDI systems. This component aims to capture snapshots and generate logs of the multi-agent system implemented using the Jacamo
framework. The log file can be uploaded in the Explanation Component to explain the agent behaviour.
The primary levels of abstraction identified in this study are as follows:
- The Implementation Level: a detailed narrative closely related to the Agent and Multi-Agent Systems for debugging and testing.
- The Design Level: a level that describes the agent behaviour regardless of its implementation for analysis and verification.
- The Domain Level: a high-level narrative that explains the behaviour with reference to the domain and requirements of the system - (work in progress).
Please use this version of Jason or JaCaMo in your multi-agent system project.
For Jason applications:
dependencies {
implementation 'io.github.jason-lang:jason-interpreter:3.2.1-SNAPSHOT'
}
For JaCaMo applications:
dependencies {
implementation 'org.jacamo:jacamo:1.2-SNAPSHOT'
}
You can follow the configuration instructions below to use the logging component as a library in your project.
- Add this library as a dependency in your project gradle file:
dependencies {
implementation 'io.github.yan-elena:agent-logging:latest'
}
- In your JaCaMo project, configure the
ag-arch
andag-class
in your.jcm
file as shown below for each agent you are interested in logging.
agent bob {
ag-arch: log.LoggerArch
ag-class: log.LoggerAg
}
If you want to use it in a Jason project, follow this configuration in your .mas2j
file:
agents:
bob agentArchClass log.LoggerArch
agentClass log.LoggerAg;
- Launch the application! The agent's logs generated as readable files (
.log
) and as.json
files are located in the/log
folder. - You can try our web application, to explore the narrative of the system at multiple levels!
An example of the configuration and use of this framework can be found in this repository: Domestic Robot Example