This site for evaluating the spread of junk news on Facebook was built by the Computational Propaganda project (COMPROP) at the Oxford Internet Institute (OII), University of Oxford. Our aim in building this website and its suite of tools is to help researchers, policy makers, and social media platforms understand the impact of junk news on public life. We do not endorse the spread of junk news, and do not make money by aggregating this content.
This work is funded by the European Research Council through the grant “Computational Propaganda: Investigating the Impact of Algorithms and Bots on Political Discourse in Europe,” Proposal 648311, 2015-2020, Philip N. Howard, Principal Investigator, and the grant "Restoring Trust in Social Media Civic Engagement," Proposal Number: 767454, 2017-2018, Philip N. Howard, Principal Investigator. We are grateful for additional support from the Open Society Foundation and Ford Foundation. Project activities were approved by the University of Oxford’s Research Ethics Committee, CUREC OII C1A 15 044, C1A 17 054. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the researchers and do not necessarily reflect the views of the funders, or the University of Oxford.
The Junk News Aggregator was built by Dimitra (Mimie) Liotsiou. Thanks to Bence Kollanyi for his valuable contributions to the data collection. Thanks also to Lisa-Maria Neudert and Vidya Narayanan, and to the rest of the COMPROP team, for their very helpful suggestions, feedback and support during the development of this project.
Many thanks to Shaun A. Noordin, and to Mike Antcliffe of Achromatic Security, for their valuable web development and security testing work on this project.
October 2019: The Top 10 List and the Visual Grid now display the top most engaged-with junk news posts over the whole data collection period. They previously displayed the top posts from the last 24 hours. The main Explorer app now features a calendar widget allowing users select posts over a desired period.
September 2019: The public Facebook posts displayed on this site were last updated on 4 September 2019. From 4 September 2019, our access to the public Facebook Graph API has been suspended, so further data has not been colleced since.
At the top of all three aggregator tools, you can choose which region you want to track. At the moment, you can choose between the EU (default) and the US.
Using the main Explorer
The interactive Explorer shows posts uploaded by junk news sources onto their public Facebook Page. It collects posts on an hourly basis, and goes back up to three months.
In addition to selecting which region you want to track, you can also select the language of the sources you'd like to track ("Language" drop-down list).
You can select the time period of interest by specifying start and end dates ("Showing Facebook posts from ... to ... ").
You can also optionally filter results by keyword (in the "optionally filtered by content" box) or by publisher name.
In addition, you can sort by Newest or Oldest posts in the selected period, and/or by each engagement type, e.g. to get most Liked posts first. And/or you can sort by the age-adjusted version of the latter.
Once you have selected all options and optionally entered any keywords, click the "Search »" button.
The age-adjusted engagement metrics offer a better way for comparing engagement levels across posts, as, due to the Facebook Graph API’s rate limits, not all posts are collected at the exact same instant. E.g. one post collected 1 minute after it was posted, another collected 43 minutes after it was posted. At most, a post may have been collected an hour after it was posted — our data collection fetches new posts from Facebook every hour on the hour. So, to adjust for this difference in post age at the time when post data was collected, we have these age-adjusted engagement numbers (they equal the number of engagements divided by the post’s age in seconds).
Using the Top 10 List
The Top 10 List is a simpler version of the main Explorer. It is not interactive, and shows the top 10 most engaged-with junk news posts over the whole data collection period (30 October 2018 to 4 September 2019 for the US region, and 20 May 2019 to 4 September 2019 for the EU region), as described in the methodology.
Using the Visual Grid
The interactive Visual Grid serves as an image-based top 256 visual aggregator tool. It shows the 256 most engaged-with junk news posts over the whole data collection period (30 October 2018 to 4 September 2019 for the US region, and 20 May 2019 to 4 September 2019 for the EU region), as explained in the methodology. Each image in this grid corresponds to a junk news Facebook post. Hovering over an image reveals a pop-up showing more information about the relevant post: the Facebook Page that posted it, the time and date posted, the text in the post, and engagement numbers. Clicking on an image takes you to the Explorer, where you can explore junk news in greater detail.
The Junk News Aggregator and its tools identify "junk news" sites based on a grounded typology tested in several scientific studies conducted by the COMPROP research group. It reflects our scientific opinion and evaluation, and does not necessarily reflect the views of our funders or the University of Oxford. The basis of opinion is described in our methodology.
For any enquiries or feedback related to the Junk News Aggregator, please email: email@example.com.