Ev·er·green (adj.): universally and continually relevant; not limited in applicability to a particular event or date.Merriam-Webster.com Dictionary
For the past few decades, the proliferation of disinformation has increased at an alarming rate, which makes understanding and reacting to pivotal events, like election cycles and public health crises, more difficult. In response to this, many governments have developed or are developing policies to identify and counter disinformation.
Unfortunately, researchers and citizens can only investigate these policies by scanning web pages and online documents, then comparing findings between relevant countries and their policies. This process takes a long time and is plagued by obstacles like the differences in policy making systems and language difficulties. Project Evergreen addresses this inefficiency by creating the world’s largest and most comprehensive database of policies pertaining to disinformation.
Formed in the Spring of 2022, Project Evergreen and the Global Disinformation Policy Database (GDPD), live at The University of Texas at Austin. The Project team working on the GDPD is split between the graduate students working on their final graduate project, the Policy Research Project (PRP), and undergraduate students who serve as research assistants within the Global Disinformation Lab (GDIL).
Undergraduate research assistants sift through thousands of links to potentially relevant websites, looking for pertinent policy “leads”. Once a lead is confirmed, it is handed off to the PRP, who then spend hours qualitatively coding the policy, as well as filtering out duplicates or mistakes. If a policy makes it past both the undergraduate researchers and the PRP team, then it is added to the GDPD.
The bisection of the project was by design; it allows individuals to join the project at any stage of their knowledge, interest, or experience – hence, the project name “Evergreen”. Those with more experience in identifying and deciphering policies join coding efforts. Novices to policy and disinformation serve as the first filter by discovering leads. This functional distinction between the two halves is not without issue; it can be difficult to receive feedback exclusively through a Slack channel. However, issues at the beginning are quickly resolved as the students fall into a routine, becoming faster and more accurate in their respective tasks.
Still, developing a comprehensive collection of disinformation policies requires searching every corner of the internet. Rather than utilizing inefficient human labor to do so, Evergreen employs a search automation service, SerpApi, to collect tens of thousands of search results within hours. A SERP, is a “search engine results page”. For example, googling for the nearest coffee shop results in a SERP listing links to web pages related to “coffee shops”, and your location. Examples of relevant keywords for this project include “disinformation”, “policy”, and a country name. An API, or “application programming interface”, is a method of computer software communication. Using an API call, the data from the SERP is transferred to a spreadsheet, which Evergreen research assistants then filter by hand.
The SERP data is divided into geographical regions: Africa, Asia, Europe, Oceania and the Americas. Categorizing the data in this way provides a deeper understanding of the cultural and political landscape specific to each region. This also provides Evergreen researchers an opportunity to specialize in a region and familiarize themselves with distinct data patterns.
The completed Global Disinformation Policy Database will enable policymakers to access qualitative and quantitative data to better understand the disinformation policies of their allies and rivals alike. Journalists will be able to contextualize and make comprehensive comparisons for their stories. Academic researchers will be able to understand the intricacies of how various governments define and respond to disinformation.
The database will be freely accessible starting in May at disinfo-policy.org. Sign up for GDIL’s newsletter to receive updates and be notified when the database is available. Read our previous blogs.