Belief Revision engine for the second assignment in Introduction to Artificial Intelligence course at DTU 02180 May 2022

With this project it is possible to create a belief base, query it for entailment, and to revise it.

Using the agent, there are two possible ways to populate the belief base. The first one is to populate a belief base with a pre-defined example from exercise 9.1 from the exercises. The other way is to define each clause manually through expansion.

Revision is then done through further expansion, or contraction. When expanding the belief base, the negated input formula is first contracted from the belief base, before it is added.

To run the program simply run menu.py by typing py menu.py in a terminal. Several tests that check AGM postulates have also been made, these can be run by running agm_postulates.py by typing py agm_postulates.py in a terminal.

When running the menu file, then a menu with 6 options will be shown. In order to choose an option, type the number in front of the desired option into the terminal.

Each option is described in more detail under menu options. For option 2-4, after choosing the option, you have to enter a propositional logic formula.

The syntax for a propositional logic formula can be seen under Formula Syntax.

1. 'See belief base': This option will simply print the current belief base into the console
2. 'Query knowledge base for entailment': Enter a formula that will be queried with the existing belief base
3. 'Expand Belief Base': Enter a formula that should be added to the belief base through expansion
4. 'Contract Belief Base': Enter a formula that should be contracted from the belief base
5. 'Reset Belief Base': Delete all clauses in the belief base.
6. 'Exit': Terminate the program

Syntax for propositional logic sentences:

• Available variables are all letters from A-Z.
• Available operators are (the syntax is written in parentheses): not(~), and(&), or(|), implies(>>), if-and-only-if(%) Example: (A&B)>>(E|D)