Skip to content

Latest commit

 

History

History
executable file
·
125 lines (87 loc) · 5.11 KB

README.md

File metadata and controls

executable file
·
125 lines (87 loc) · 5.11 KB
Raw

ChatbotSQL

Introduction

ChatbotSQL is an educational platform to learn SQL by solving exercises. This app will be able to help students thanks to our virtual assistant, created with IBM Watson.

With this platform, we try to improve the learning experience of Databases subject students at University of Cádiz, at the same time that we analyze the most common mistakes and questions.

Description

Main Menu:

Once we log in, we can choose between doing exercises or trying the playground mode.

  • The exercises will test the user's skills. There are three difficulty levels, which will be unlocked when we solve all the exercises from the previous level.
  • The playground mode will allow users to perform some queries freely.

Image text

Game tab:

When we select an exercise or the playground mode, we will see three panels:

  • On the left, we can see the data stored in the database without performing a query. We can also select another problem, and check the students that have completed the current exercise.

  • In the middle, we have the text area, in which we have to write the query. The result of the query will appear here too.

  • On the right, we have our virtual assistant. It will give us some information.

Image text

Contact:

In this tab we can report the errors we may find. We can also rate the app.

Image text

Running locally

Compilation requirements, operating environments & dependencies:

  • PHP v7.4
  • Composer v2.1.9
  • Node v16.13.1
  • NPM v8.1.2
  • MySQL 8.0
  • Apache2 Server
  • IBM Cloud account

Import this repo:

git clone https://github.com/rubenperezm/ChatbotSQL.git

Create 3 databases in MySQL and run the scripts included in misc/data. ChatbotEN.sql includes some exercises, db1EN.sql and db2EN.SQL are the databases used in the exercises.

App

Go to src/App and run these commands:

sudo apt-get install -y php7.4-{xml,bcmath,gd,mbstring,xsl,zip,curl,mysql}
composer update
cp .env.example .env

We created a .env file. Edit this file to include the databases previously created. You can also modify the URL of the platform, and the URL of the chatbot if needed.

Now, run these commands:

php artisan key:generate
sudo chown -R $USER:www-data storage/
sudo chown -R $USER:www-data bootstrap/cache/
sudo chmod -R 775 storage/
sudo chmod -R 775 bootstrap/cache/
php artisan ser

At this point, the platform should be visible at http://localhost:8000

IBM Watson Assistant:

Log in to IBM Cloud and create an IBM Watson Assistant service instance. Import the assistant included in misc/IBMWatson.

Intents

Intents are purposes or goals that are expressed in a user's input. The most important intents are:

  • Help: The user asks for generic help
  • Hints: The students asks for a hint
  • SyntaxError: Error code generated by the platform when the user's SQL query returned an error code.
  • WrongQuery: Hint generated by the platform to help the student after running a query that returned a different result to the solution query.

Entities

Entities are used for identifying interesting parts of the user's utterance, such as clauses and functions. Watson Assistantcan recognise these keywords and all the different ways in which the user can refer to them. The most importants entities in our instance are:

  • Clause: Includes the most common clauses in SQL ('SELECT', 'FROM', ...), functions ('AVG', 'MAX') and all the different ways in which the user can refer to them.
  • DBJargon: Common terminology used in Databases ('Primary key', 'schema', ...).
  • ErrorCode: The error codes of the most common MySQL errors.

Dialog

A dialog skill uses Watson natural language processing and machine learning technologies to understand user questions and requests, and respond to them with answers that are authored by you. The message the skill returns depends on the intent and entities indetified in the user's message.

Image text

BotBD

BotBD is a modified version of assistant simple.

Go to src/botBD, and create a .env file:

cp .env.example .env

Now you need to complete ASSISTANT_URL, ASSISTANT_ID and ASSISTANT_IAM_APIKEY fields with the information of your assistant. You can find this information in your assistant settings.

Before you run the server, please change the first line of botBD/public/js/conversation.js to include the current URL of the Laravel platform.

Once you have done this, run these commands to install all the dependencies and run the NodeJS server:

npm install
npm audit fix
npm start &

The chatbot will be running on port 3000.

About the project funding

This research & development is part of the project CRÊPES (ref. PID2020-115844RB-I00), funded by MCIN/AEI/10.13039/501100011033/.

It is also partially granted by the EMERGIA programme, funded by the Junta de Andalucía through the grant EMC21_004171.

This project has been also funded in the call of Teaching Innovation of the University of Cádiz 2021/22 “Proyecto de Innovación Docente de la UCA" (code sol-202100203360-tra).