Newsworthy is an app that draws on artificial intelligence to assemble news articles that match individual users’ interests. The idea is to prioritise actual content over authorship.

Combining Curation and Social Networks for Better-Targeted News Content

Could this be the new Spotify for news clips? This at least is how Australian entrepreneur Mehran Granfar, who is currently running a campaign on the Kickstarter crowdfunding platform, is touting his Newsworthy app. The app provides a curation service for new items to decide which clips should be targeted at each individual user. As with Spotify in the music sphere, the system uses artificial intelligence (AR) to analyse a Newsworthy user’s main areas of interest and offer information most likely to interest him/her. However, the service goes somewhat further than what is done by traditional curators. In addition to using key words corresponding to particular subjects, Newsworthy adds automated tracking of the published views of opinion leaders on a given topic. Along the lines of Twitter, where you can follow Tweeters with similar interests to yourself, Newsworthy will enable you to find relevant information based on assessments by other commentators whom you have decided to follow.

Shaking up online ‘sharing’ habits

To appreciate what differentiates Newsworthy from other curation systems and social network sites, you need to understand that the app aims to prioritise actual content over the people posting it. Newsworthy allows you to distinguish ‘thought leaders’ from the people you know in real life who habitually make up our social networks on Facebook and LinkedIn. The novel feature here lies in the practice of following an opinion leader in order to find quality articles that s/he rates highly. Mehran Granfar wants to overturn the current social network paradigm which dictates that an article is worth reading if it is ‘liked’, ‘shared’ and ‘commented’ more than others. Newsworthy draws on implicit, automated sharing of content that the app deems interesting for you, the user, based on an actual reading of articles by respected commentators.

Voting helps to hone recommendations

The AR-based app is designed to learn from our reading experiences as we use it and gradually hone its offerings through a user-friendly voting system. Simple left and right swipe gestures allow you to vote an article either up or down in order to rate its newsworthiness. Those you approve of will be served up to your followers, and the app will also endeavour to find similar articles to serve up to you. Mehran Granfar argues that this kind of voting is part and parcel of the experience and that this technique is the best way of providing high-quality, well-targeted curation, so voting should not be seen as a chore. If Granfar hits his funding target he is planning a beta launch of the app for iOS and Android in September or October, and a Windows version will then be developed if the service proves popular. Granfar also envisages creating a more comprehensive service on a web platform.

By Lucie Frontière