Johan Bollen, an associate professor of informatics and computing at the School of Informatics, Computing, and Engineering, is tackling media independence and bias as part of a paper that will be published in the latest edition of the journal Public Library of Science ONE.
Bollen co-authored the paper with Erick Elejalde, a former Ph.D. student of Bollen’s who is now working at the Data Science Institute of the Universidad del Desarollo in Santiago, Elejalde’s local Ph.D. advisor Leo Ferres, and two other Chilean colleagues. The group used social media data and ownership records to study how the consolidation of Chilean media outlets has impacted the diversity of stories reported in the country.
“During an extended visit to Chile in 2015, we all become particularly interested in whether a data-driven analysis could shed light on long-standing issues in journalism,” Bollen said. “Namely, whether the media addresses communities equally, whether certain topics are preferentially covered or not, and whether there were any relationship between these issues and the ownership structure of media.”
Bollen and his colleagues discovered that traditional and social media have become intertwined, allowing the distribution of news to expand outside of local regions to become influential in shaping a variety of social phenomena. A lack of clarity of the ownership of traditional media has blurred coverage, with a handful of entities owning a variety of outlets that allow a wide range of stories to seemingly come from multiple sources when, in reality, they all come from a relatively small group of owners.
“Chilean media is highly concentrated both in terms of ownership as well as in terms of topics covered,” Bollen said. “Moreover, ownership seems to play an important role in news content similarity. In other words, who owns the outlet seems to have an effect on what news gets reported and how. This may reduce diversity and coverage.”
The methods described in the paper “Power Structure in Chilean News Media” will allow researchers to analyze large-scale, language-independent social media data to identify posts that feature the same style and information. Such analysis can can lead to the discovery of the origin of news reports.
“Transparency is key,” Bollen said “Most traditional newspapers are known to cover the news, even if perfectly objective, from a particular angle, and are part of known groups and ownerships. Their readers are aware of this and can account for it when they consume the news.
“This might not always be the case on social media like Twitter, which is designed to be very brief in content yet massive in scale. This may overwhelm the capacity of individuals to make an informed reading of the news, in terms of its veracity, its origin, and its importance.”
Bollen and his colleagues’ research focuses on Chile, but they hope to expand it to other countries to analyze media bias and the relationship between commercial interests, ownership, and the public.
“We hope our research can lead to computational tools that can provide more transparency to users about where their news comes from, why certain topics are being covered, and who specifically cares about whether something is fit to be covered as news or not,” Bollen says.
For more information on complex networks and systems research at SICE, visit our website.