The Problem

Today’s media interfaces present social opinions to us with a monotone composition. The lack of balance and diversity in content has created filter bubbles all around us, leading to a loss of common understanding. Admittedly, getting our news from various social media sources is much better than relying on one or two TV networks or newspapers but their relevance based algorithmic approach isn’t cutting it. Burst Your Bubble is a web application that attempts to broaden your political awareness!

Burst Your Bubble! An Intelligent System for Improving Awareness of Diverse Social Opinions

(This paper is the foundation of our comps project and we used it as a jumping off point for our work. While the problem we decided to tackle was the same as the paper, our approach as differed greatly)

Abstract: Social media users are overloaded with diverse opinions by people with opposing stances. Previous research shows that people often look for opinions that reinforce their pre-existing beliefs and stances, which may lead to social polarization. Traditional social media present opinions in a linear list format, which not only lacks structures for people to explore diverse viewpoints but also aggravates their selective exposure to agreeable opinions. To address this problem, we designed an intelligent system that improves awareness of diverse social opinions by providing visual hints and recommendations of opinions (e.g. news articles and comments) on different sides with different indicators. We evaluated our system with news articles about Obamacare repeal issue and their corresponding user comments from Facebook. Results demonstrate that our system could increase people’s awareness of their stances and opinion selection preferences, which mitigates selective exposure and thereby leads to a more balanced perception of social opinions.

Read Paper

The Web App

Our web application delivers a personalized news feed to you by gathering the latest and most important topics of the day, learning about your political biases on them, determining the biases presented in various articles, and presenting a balanced distribution of articles to keep you best informed.

See App

The Product

Design

The design process for this project was extremely iterative. We used a classic HCI design cycle model to inform our process and took the project from discovery phase to evaluation.

Learn more

Stack and Infrastructure

We used a variety of technologies to build BYB, including frameworks we had familarity with, and some that we wanted to learn!

Learn more

News Classification

Utilizing machine learning to determine the political stance of online news articles.

Learn more

User and Article Models

Modeling the user behaviour with a User Score and updating Article stance on user feedback

Learn more