NLP Logic Inference Engine

Introduction

This project aimed to perform logical reasoning on plaintext English statements. Allowing for automated reasoning, in which the program relies on an internal knowledge bank. It also includes a manual mode, where the program allows you to define your own rules and input statements you would like those rules to be evaluated on.

Building the application

Requirements

dotnet SDK 2.0 or higher

node.js 6.11 or higher

A JS package manager: yarn or npm

Automatic

The main section of the application is located in the .\WorldBuilder folder. Here you will find 4 batch files.

For a release build run:

setup.bat

release.bat

For a development environment run:

setup.bat

development.bat

then in a separate terminal:

run.bat

Purpose of each batch file

setup.bat : Must be run at least once, to install Javascript and Node dependencies. Ensure you have satisfied all the dependencies listed above before running this.

development.bat : To start FABLE daemon and launch a development server, use in conjunction with run.bat

run.bat : To be run after a development.bat instance has been launched, to view a reactive version of the application.

release.bat : To build release version of the application, complete with windows executable.

Manual

Building the application

Open a terminal in the .\WorldBuilder folder.

Install JS dependencies: npm install or yarn install

Install F# dependencies: dotnet restore

Start Fable daemon and Webpack: npm start or yarn start

In a separate terminal, run: npm run launch or yarn run launch

Building application for release

Open a terminal in the .\WorldBuilder folder.

Run: npm run build or yarn run build

Run: npm run release or yarn run release

You should have now a .\WorldBuilder\release\NLPLogicEngine-win32-x64 folder. Inside you will find the NLPLogicEngine.exe, which will be an executable version of the application.

Contributors

This project was a term-long group project for High Level Programming module taken in Spring 2020. I collaborated with: Neelesh Ravichandran (Neelesh99), Preet Lalli (preetl), and Agrim Manchanda (agrimmanchanda) all of whom were excellent team mates.