The Challenge
Online news outlets often suffer from a dearth of civility and quality in the discussions that emerge in news comments. Incivility can further polarize people's beliefs on issues. But there are high quality comments that are made too. Research shows that by curating and highlighting such high quality comments it can signal norms and expectations for comment discourse. This can improve the long term quality and viability of online discussion around important civic issues.The Project
The CommentIQ project is about making it easier for community managers and moderators in a range of news organizations to quickly identify high quality comments that they can highlight on their sites. We are examining a range of automatically computed scores that can help identify the good stuff from the not-so-good stuff.The project consists of four interrelated components:
Several research papers on various aspects of online comments including:
- S. Sachar, N. Diakopoulos. Changing Names in Online News Comments at the New York Times. Proc. International Conference on Web and Social Media (ICWSM). May, 2016. [PDF]
- D. Park, S. Sachar, N. Diakopoulos, and N. Elmqvist. Supporting Comment Moderators in Identifying High Quality Online News Comments. Proc. Conference on Human Factors in Computing Systems (CHI). May, 2016. [PDF] (Best Paper Honorable Mention)
- N. Diakopoulos. Picking the NYT Picks: Editorial Criteria and Automation in the Curation of Online News Comments. #ISOJ Journal. April, 2015. [PDF]
- N. Diakopoulos. The Editor’s Eye: Curation and Comment Relevance on the New York Times. Proc. Conference on Computer Supported Cooperative Work (CSCW). March, 2015. [PDF]
- J. Hullman, N. Diakopoulos, E. Momeni, E. Adar. Content, Context, and Critique: Commenting on a Data Visualization Blog. Proc. Conference on Computer Supported Cooperative Work (CSCW). March, 2015. [PDF]
An open-sourced API to compute comment quality scores including metrics for article and conversational relevance. See the Github page for a demo or to checkout the code, as well as to see the specifics of the quality scores we compute. You can integrate the data processing afforded by the API into your website backend to enable new ways of moderating comments, or to create new user-experiences.
A moderator interface to demonstrate how visual analytics using the scores can enable newsroom moderators to find higher quality comments. Try out the demo here.
An end-user interface to demonstrate the API as well as new user-experience possibilities. Check it out here.
About
CommentIQ is a product of the Computational Journalism Lab at the University of Maryland, College Park College of Journalism and is funded by the Knight Foundation Prototype Grant program.For more information, contact Nick Diakopoulos at nad@umd.edu.
