Expert System

École 42 assignment - an expert system for propositional calculus using backward-chaining algorithm

About the project

• This is a group project of École 42.
• My teammate is keatchut.
• The objective of this project is to create a program that can take logical statements, their initial value and use backward chaining algorithm to find the value of the queried facts.

System Requirements

• Compiler that support C++17
• CMake 3.10+
• Boost library 1.68+

Installation

Requirements

• MacOS

  brew install cmake
  brew install boost
  

• Ubuntu 18.04

  sudo apt update
  sudo apt install cmake
  sudo apt install libboost-all-dev
  

Expert System

mkdir build
cd build
cmake ..
cmake -build .
cd ..

The executable will be at bin/expert_system.

Usage

To run in normal mode

./bin/expert_system [filename]

Options

Shorthand	Full	Mode	Description
-h	--help		show help message
-f	--facts	interactive facts mode	user can change initial facts without having to rerun the program
-q	--query	interactive query mode	user can change query without having to rerun the program
-v	--visualisation	reasoning visualisation mode	display step-by-step backward chaining reasoning process
-d	--debug	debug mode	display all values after evaluation
-c	--color	color mode	display with color
-p	--print	print mode	print input file

Input

Below is an example of an input.

A + B => C
C => D

=AB

?D

The input are seperated into 3 sequential parts:

• rules

  Format: condition => conclusion.

  A condition contains one or more facts and operators on the list below (ordered by priority of execution).

  • NOT !
  • AND +
  • OR |
  • XOR ^

  A conclusion has one or more facts and can have only operator AND +

• initial facts

  Format: =[facts]

  All fact values are default to FALSE. The initial facts declaration will change the fact values to TRUE.

• query

  Format: ?[facts] The program will do backward chaining on the facts and print the result.