STEP (Symbolic and Tractable Explanation Package)
Generates explanations from a Sentential Decision Diagram (SDD). The explanation techniques are described in the following paper:
A Symbolic Approach to Explaining Bayesian Network
In Proceedings of the 27th International Joint Conference on Artificial Intelligence (IJCAI), 2018.
Andy Shih and Arthur Choi and Adnan Darwiche.
Input Format
The input SDD is specified using "config.json". It should have a JSON object with the following fields:
- "name": the name of the SDD of interest
- "sdd_filename": the filepath for the SDD
- "vtree_filename": the filepath for the vtree accompanying the SDD
- "variable_description_filename": the filepath of the variable description file for the SDD
First, each occurrence of %s in the configuration paths will be replaced with config["name"].
Then, the SDD file should be located at config["sdd_filename"],
the vtree file should be located at config["vtree_filename"],
the variable description file should be located at config["variable_description_filename"].
For example, if config["name"] = "admission_1" and config["sdd_filename"] = "sdd/%s/%s.sdd" then the SDD file should be located at "sdd/admission_1/admission_1.sdd".
Output Format
The explanation output will be printed to stdout.
Running BNC_SDD
Edit src/expl.py to play around with the different explanation queries available in src/explanation_queies.py.
Then, run
./run
Further Questions
Contact us at
Andy Shih: andyshih@cs.ucla.edu
Arthur Choi: aychoi@cs.ucla.edu
Adnan Darwiche: darwiche@cs.ucla.edu