Honours Project 2018 - 2019

A chatbot which can answer C++ questions based on data collected from Stack Overflow.

Features

• Command line inteface for easier data collection, training, and deleting of files.
• Alternate responses.
• Rating responses.
• Responsive design.

Demo

There is a live demo of the project at https://ligh7bringer.pythonanywhere.com/.

Repo Status

Branch	CI	Code coverage	Requirements	Maintainability
master
development

Prerequisites

• Python 3.6 or newer
• pip

Setup

1. Clone this repo and cd Chatbot/
2. pip install -r requirements.txt (you may want to create a virtual environment before running this command)
3. Environment variables required by Flask are set automatically (thanks to .flaskenv)

Commands

After installing the required libraries, your options are:

• flask --help - shows the available commands
• flask crawl - collects data from Stack Overflow which can be later used for training the bot (stored in chatbot/training_data/). Optional command line arguments are:
  • -t, --threads <number of threads to be used>
  • -p, --pages <number of pages to be crawled by each thread>
  • -v, --verbose - verbose command line output
• flask clean - deletes the data collected with flask crawl (stored in chatbot/training_data/)
  • -y, --yes - don't ask for confirmation
• flask train - requires data collected from the crawl command and trains the chatbot using it
• flask del_db - deletes the database generated after training the chatbot
  • -y, --yes - don't ask for confirmation
• flask run - runs a local server. The website can be accessed at 127.0.0.1:5000

Tests

Unit tests can be run from the root folder of the project with the command pytest after installing the dependencies.

Screenshots