Radio France: "public service algorithm" launched

Radio France has launched a "public service algorithm" for its platform, the company announced on Wednesday, November 23. It is a hybrid recommendation method : algorithm and human.

It was developed following five main principles (human, transparency, discovery, efficiency and control) "to respect the values of Radio France," according to the press release. "Our role as a modern public service media is to work to make digital more ethical," Sibyle Veil, CEO of Radio France said in the press release. The operation of this algorithm is transparent to users.

In concrete terms, the editorial team chooses the podcasts which are displayed in the app's carousel selections or on the radiofrance.fr site. If the personalization has been accepted by the user (it can be modified at any time), the algorithm then intervenes and the order of the podcasts is then rearranged and personalized according to the user's past listening, thanks to new indicators discoverability, which avoids the lock-in bubbles produced by traditional algorithms. Publishers always have the last word in the choice of content highlighted and can adjust the recommendations.