Implementation of Shallow Neural Network to Detect Political News on Web App Bold

Overview

BOLD helps manage election hoax news by predicting the veracity of news articles and offering educational resources on news verification and combating hoaxes. Users can also report suspected hoaxes and provide feedback on the prediction model's accuracy.

Background

Between 2020 and 2021, the spread of hoaxes increased from 11.2% to 11.9%. The most common type of hoax news was about health. In 2019, which was an election year, the number of hoax news reached its highest point with 928 cases. In 2022, the number of hoax news rose again. It is expected that more and more hoax news will appear before the 2024 elections. Hoax news can cause problems like threatening the quality of democratic events, damaging voters' ability to think clearly, making the election process seem less valid, and harming community harmony, which can lead to the country falling apart. Because of these impacts, it is important to deal with hoax news.

Methodology

Our team comprises technology enthusiasts with expertise in Data and AI, and Full Stack Development.

• The Data and AI division is responsible for developing classification algorithms to predict the validity of news.
• The Back End division is accountable for constructing endpoint APIs, establishing web application infrastructure, and utilizing Firebase hosting for web application deployment.
• The Front End division is responsible for designing the web application interface using Figma, and subsequently implementing the design using HTML, CSS, and ReactJS.