Crowdsourcing-Ranking-of-Tweets

This project is based on my Final IS Project. It entails the collection and labelling of tweet by social media users in order to identify misinformation and provide datasets for automating the process using machine learning

Modules of the system

The project consists of two modules: The Pybossa Server and the Streamlit application.

Pybossa Server

This is the main server that manages the crowdsourcing process from creating a project, labelling and API to send results to the Streamlit server. This module allows the user to download the task results as well for independent data analysis.

Streamlit Server

This is an open source framework was used to create the interactive results dashboard. The dashboard gets the labelled tasks from the Pybossa API known as Enki.

Note: For more details on how to install them, please find documentation links below on how to get started.

Pybossa Server Installation links:

1. Getting started
2. Installation
3. Configuration
4. Using command line - used to install the project

Deployment links:

1. Deployment Process
   • Look at Pybossa README for more details

API for Pybossa

1. Learn about the Enki API and how to install it - this was used in the Streamlit application
2. Learn about the webhook application for real-time data analysis
   • This is a different method from using Streamlit. However, it is not operational
3. Official Documentation for webhooks - documentation on webhooks for real-time analysis

Streamlit Documentation

1. Learn about Streamlit