NewsQ design thinking activities are focused on three main outcomes: (1) technical definitions, (2)“ideation” through data, and (3) creative prototypes.
In our mission to improve news ranking and recommendation systems, NewsQ asks journalists, researchers and technologists to use journalistic standards and civic frameworks to clarify how news should be defined and framed for online realms.
Since algorithms operate according to defined rules, NewsQ documents technical definitions around the principles and metrics of online news quality that can be more readily incorporated into algorithmic processing. Our current activities work on:
➔ Methodological definitions and metrics
Methodological definitions and metrics are proposed in essays and papers, ranging from short, reflective blog pieces that clarify the inquiry around news quality for machine purposes to structured research designs about metrics that can help determine how to assess the state of online news.
As our activities help bring clarity to these terms, we’ll share those insights here. In addition, during 2020, our “project of the year” is to better define, or even attempt to quantify, news for online contexts.
- Social Science Research Council NewsQ Call for Papers – News Quality in the Platform Era
➔ Potential news quality “signal” or indicator data definitions
As of 2020, the NewsQ Database contains over 100 units of information on over 12,000 news and information sites. Some of these units of information may be potential “signals” of quality journalism.
Signals, or potential indicators of information quality or credibility, can include a number of content- and context-related information, such as:
- the emotional valence of content, or
- the author of an article.
When taken in context in combination with other signals or information, these indicators may help point algorithms to higher quality journalism. For more thoughts about what a signal is, W3C Credible Web Community Group’s report Technological Approaches to Improving Credibility Assessment on the Web is a good place to start.
- Signal categories and classes. We strive to collect seven classes of signal information, which you can read more about here.
- Signal definitions. A partial draft of signal definitions, according to certain categories, can be found described here.
Ideation through Data
At the same time, NewsQ uses our data collections to “ideate” or foster new ideas that help clarify what constitutes – and how to define – quality in news ranking and recommendation.
We collect two types of data to create opportunities for insight:
- A database, described above, of signal data across 12,000+ information outlets
- Archived screen captures of news ranking and recommendations from various platforms and applications
This data is used to foster new ideas through:
A pilot of NewsQ’s database of signal data has been offered to invited researchers and analysts in order to inform their work on news quality; parameters for broader access are under development.
➔Panel Discussions among Journalists
An upcoming series of activities will bring together journalists and technologists to review the archived captures in order to develop practical guidelines for understanding what constitutes quality news online.
Our first panel review process will take place from May – August 2020.
- News Ranking Review Panels description
NewsQ also facilitates solution prototypes to the news ranking and recommendation challenge among people with different interests and backgrounds.
A final activity area for the project involves creating an ecosystem of cooperation that centers on advancing insights into the data. Examples of NewsQ creative prototyping activities include:
➔Seed Grants for Data Projects
NewsQ has offered a few grants for data projects that may surface journalistic standards for news ranking and recommendation systems. These projects will also eventually give data back to the NewsQ Database. One example recipient is the Markkula Center for Applied Ethics, and their project on journalistic ethics and behavior:
Grantees also participate in collaborative events where they share their progress and have the opportunity for critical feedback.
➔“NewsRank for Social Good” Hackathons
Building news ranking and recommendation systems can be tough. Their results make their way into a number of products, but are most easily seen in News tabs or applications such as Google News, Bing News, and Apple News and their lists of recommended news articles.
What factors determine which articles get top billing? And how should social values be incorporated into the mix? Using a hackathon format, we are inviting students to offer their creative approaches to the problem.