Personalised news: Perspectives of journalists and users

The research team led by Frank Esser (University of Zurich) is investigating the effects of news recommender systems on the media landscape and users as part of the National Research Programme NRP 77.

News recommender systems (NRS) utilise algorithms to tailor content suggestions on social media platforms and online news portals based on users' past interactions. These systems not only consider an individual's previous reading patterns but also align recommendations with the interests of similar users.

What role do these systems play for media companies in Switzerland and the Netherlands and how are these systems perceived by users in an international comparison? The research team led by Frank Esser (University of Zurich) answers these questions as part of NRP 77. To do so, they first surveyed 36 experts working in media companies and then conducted a representative survey with users in five countries.

Media professionals are sceptical about NRS

In Switzerland, news recommender systems are still in the testing phase. Media companies are experimenting and still looking for the optimal use of these systems on their news portals. Although the experts recognise that NRS can provide readers with orientation in the thicket of articles, there is a concern that editorial control will be weakened by the automated recommendations. According to the research results, many media professionals also doubt the concrete benefits of NRS.

Conclusion of the researchers:

“with the common goal of striking a balance between algorithmic personalisation and editorial integrity, the development of NRS can move towards “responsible“ NRS that consider transparency, user control, and diversity. Ultimately, introducing NRS can also lead to new working practices and forms of collaboration that can positively impact news work.”

Positive attitude among users towards algorithmic news recommendations

The readers surveyed generally show a positive attitude towards algorithmic news recommender systems. They see the benefits mostly in the areas of sports, entertainment, and celebrity news. Readers assign lower importance to personalised recommendations on the homepage and demonstrate, according to the research findings, even less interest in receiving personalised newsletters and push notifications on their mobile devices.

However, respondents said that their trust in media companies that they suspect use news recommender systems is lower.

A 'responsible' use of NRS is also what the users of the five countries surveyed by the researchers would like to see. They expect transparency about the processing of their data and disclosure when content is recommended by algorithms.

In Switzerland, users are (still) sceptical about news recommender systems. This is less true of the US and the UK, where media companies already use NRS on a larger scale. According to the survey results, users there have a more positive attitude towards these systems.

In summary, the most important points from the study include:

  • NRS can positively impact media companies and users if the necessary resources are available and technology demands are balanced with journalistic demands.
  • Institutionalised forms of collaboration and regular exchanges between IT, editorial, and business can prevent tensions.
  • To ensure that trust in the media is not compromised, it is essential that NRS are aligned with journalistic criteria (such as relevance and diversity) and that transparency and information about algorithms are provided.
  • To increase the acceptance of NRS, the benefits for users must be actively communicated, and fears about filter bubbles and data misuse must be avoided.
  • The preferences and attitudes of Dutch and Swiss media users are in many ways similar to those of users in other countries. However, Dutch and Swiss media users are more critical than users in countries where the use NRS is more advanced (such as in the US or the UK).

More detailed information on the study results can be found in the newly published factsheet:

The role of news recommender systems in digital democracies (PDF)

This research work was carried out the framework of the National Research Programme (NRP 77), which supports 46 projects related to digital transformation.