Background
SFI MediaFutures is a leading research center located in Bergen’s Media City in Norway. It focuses on creating responsible, AI-driven media technology to support ethical and effective user engagement, content production, and audience insights, in partnership with top national and international media and research organizations. Researchers at MediaFutures shared with us how NewsCatcher is supporting their work.
Before NewsCatcher
In the past, MediaFutures relied on outdated news articles which they received from their partners, making it difficult to simulate an authentic news site experience for users. Receiving real-time news feeds would require building their own scrapers—a time-consuming and costly approach—or adopting a third-party solution. MediaFutures ultimately adopted NewsCatcher, enabling the team to focus on enhancing the overall user experience.
“I used the Washington Post corpus for my experiments before, but the news was really old and outdated, which affected my results. Since the articles were outdated, this introduced an external factor that could affect my results. I measure user selection behaviour under specific experimental conditions, so I needed fresh and relevant news to ensure that my findings weren’t influenced by outdated content.” — Khadiga Seddik, PhD Candidate
How NewsCatcher Supports MediaFutures
MediaFutures regularly conducts experiments for research papers and innovation projects, striving to create user scenarios that mirror real news environments. By integrating NewsCatcher’s fresh and dynamic feeds participants receive current, authentic content, leading to more accurate data on user engagement and behavior. NewsCatcher’s automated setup lets researchers gather thousands of articles simply by adjusting parameters, substantially reducing overhead and making large-scale studies more feasible.
“Using NewsCatcher to gather the information we need is, I would say around 90% more efficient than manual collection methods.” — Tobias Wessel, Research Assistant
Currently, researchers within one specialized work package use NewsCatcher regularly. However, there is growing interest among other teams to incorporate fresh data for studies focused on online news consumption. In the long term, MediaFutures plans to make near real-time updates a benchmark for these projects, broadening the scope of experiments benefiting from up-to-date content.
“We have rarely experienced it [our previous solution] to be as up-to-date as the articles we retrieve from NewsCatcher, which provide data from the actual day.” — Tobias Wessel, Research Assistant
Example Study Using NewsCatcher
One of MediaFutures’ studies aims to mitigate selective exposure, a tendency where individuals prefer information that aligns with their preexisting views while avoiding conflicting perspectives—for instance, a Democrat primarily consuming Democrat-leaning articles. This research explores whether headline style can influence reading habits. By leveraging NewsCatcher’s near real-time feeds, participants engage with content sourced on the same day, ensuring timeliness and authenticity. Large language models then produce multiple headline variations for the same articles, and participants select their preferred style. Researchers follow up with targeted questions to understand user choices, ultimately analyzing how these behaviors can be influenced to encourage broader information discovery.
“Using NewsCatcher has been a big step forward for my experiments. I can now access fresh articles with minimum delay, which helps keep my research more realistic and accurate. It’s been a big improvement in making sure my experiments reflect actual news consumption behavior.” — Khadiga Seddik, PhD Candidate
Other Solutions MediaFutures Explored
Before implementing NewsCatcher, MediaFutures investigated multiple APIs and data providers. The team considered the GDELT API for its research-oriented features but found it difficult to implement. It also looked into solutions from Google, Bing, and Newsapi.org, though these options often did not consistently provide critical metadata—such as image URLs—or comprehensive Norwegian news coverage. Impressed by NewsCatcher’s robust documentation, case studies from well-regarded research institutions like MIT and Berkeley and responsive customer success team, the MediaFutures research team decided on the winner.
Future Developments
Looking forward, MediaFutures envisions using NewsCatcher’s real-time data to analyze emerging news trends and shifts in public discourse more effectively. Although initially limited to a few pilot projects, NewsCatcher is being integrated into an expanding range of research initiatives. Early successes have led the center to explore additional applications, further solidifying NewsCatcher’s role in advancing responsible, data-driven media research.