BALANCE: Enhancing diversity in news and opinion aggregators

Through this project, A UMSI team led by Professor of Information Paul Resnick has investigated and built diversity-aware news recommendation services to improve people's news reading experiences. One system, BALANCE, classifies news and opinion articles as conservative or liberal using data from social news aggregator Digg, which allows readers to vote for articles they like. The system combines the patterns of readers who voted that they “liked” certain articles and automatically classifies news and opinion articles as liberal or conservative. 

Another recommendation service developed through this research project, ManyAngles, offers suggestions on a wide range of topics that include technology, health, business and other areas in order to give readers well-rounded viewpoints on a myriad of subjects. ManyAngles does not apply political or social labels, rather a computer algorithm is responsible for providing suggestions to stories that it thinks are written in different perspectives.

Start date: 7/1/2009
End date: 8/31/2014

Read More

People today are given many choices of where they would like to get their news and what viewpoints they would like to read about. If readers are only exposed to articles supporting their own opinions, opportunities for opposing groups to find a common ground become more difficult. Exposure to diverse and challenging opinions can reduce the political polarization in our society and encourage civil discourse among conflicting groups.

Aggregators such as Digg, Reddit, and Google News rely on ratings and links to select and present subsets of the large quantity of news and opinion items generated each day. From this project, researchers hope to form alternative measures of diversity for result sets, develop algorithms that optimize diversity and popularity, and evaluate user interface techniques for more diverse result sets. This will help researchers to assess the impacts of alternative selection and presentation methods on people's willingness to use an aggregation service, their exposure to diverse opinions, and the size of their argument repertoires.

The development of an automatic classifier can be useful in creating news services that seek to encourage diversity in opinions. Readers will be given the opportunity to browse through articles that challenge their own views.

The ManyAngles application is currently available as a free Google Chrome extension and is intended to provide greater balance and more thorough coverage than a reader would typically receive through his or her own news consumption habits. The extension detects the topic of the article that the user is reading and produces recommendations on that topic. It collects readers’ browsing visits to only the designated news sites found on the server in order to provide more accurate suggestions and allow users to keep track of their coverage; it does not collect any personal information or outside browsing history. The results gathered after the user installs the application will allow the researchers to determine the application’s effectiveness in encouraging people to seek more diverse sources.

Ultimately, the algorithms created by the student teams can help encourage readers to view items that both support and challenge their own opinions. Readers will then be able to engage in meaningful discussions regarding their differing opinions. The results of this project will provide a better understanding of alternative notions of what it means for a set of items to be diverse, and the range of reactions that different people have to varying levels and presentations of diversity.

This project received an additional $16,000 through a National Science Foundation REU Supplement to support the work of two undergraduate students. The undergraduates were involved in analysis of comments and posts on blogs. They helped develop the coding scheme and coded a large data set that the researchers collected. While much of the coding was performed by hand, the students also explored and used automated coding of the text. Additionally, these results work may help the researchers expand their work to websites, such as forums or the comments on mass media sites.

To hear Professor Resnick talk about the BALANCE project, watch his YouTube interview:


III: Small: Optimizing News and Opinion Aggregators for Diversity, National Science Foundation: $515,312

REU Supplement: III: Small: Optimizing News and Opinion Aggregators for Diversity, National Science Foundation: $16,000

I-Corps: Diversity-aware News Recommendation, National Science Foundation: $50,000


The National Science Foundation (NSF) is an independent federal agency created by Congress in 1950 "to promote the progress of science; to advance the national health, prosperity, and welfare; to secure the national defense…"