App Functionality Proposal - Cian

 App functionality proposal.

 

Our citizen Kane app is intended to provide a balanced, unbiased source of news articles to the user.

 

Many apps already exist which filter out articles based on a users preference and world views.  While on paper this is an effective way of providing relevant news to the user, it risks placing the reader in an echo chamber, stopping them from seeing different perspectives.

 

I propose that our app takes a different path. 

Instead of filtering out articles based on preferences, our app will instead analyse news articles from a wide variety of sources and give them a score based on potential bias. It would achieve this through the use of a machine learning language model. By analysing the type of language used in an article, the app can determine the level of sensationalisation, factual accuracy, and political bias in either direction.

 

For example, analysis of articles from red top publications like The Sun or The Daily Mail would be flagged as tabloids with a high level of sensationalism and a right wing bias. On the other hand, articles from the Associated Press would be flagged as having a low level of political bias and sensationalism.

Particularly extreme examples could even be flagged as propaganda.

This would not prevent the user from viewing the article, but instead give them a view into articles they may potentially avoid otherwise.

A user feedback function could be implemented to help train the language model and improve its accuracy.

 

Allowing the user access to every publication without restriction gives them a more balanced view on news reporting. Seeing both sides of a story along with the article analysis allows the user to remain informed, and determine the most trustworthy and factual publications.