The Community Data Science Collective, in collaboration with Pushshift and others, is launching a new collaborative project to create a digital observatory for socially produced COVID-19 information. The observatory has already begun the process of collecting, and aggregating public data from multiple online communities and platforms. We are publishing reworked versions of these data in forms that are well-documented and more easily analyzable by researchers with a range of skills and computation resources. We hope that these data will facilitate analysis and interventions to improve the quality of socially produced information and public health.
During crises such as the current COVID-19 pandemic, many people turn to the Internet for information, guidance, and help. Much of what they find is socially produced through online forums, social media, and knowledge bases like Wikipedia. The quality of information in these data sources varies enormously and users of these systems may receive information that is incomplete, misleading, or even dangerous. Efforts to improve this are complicated by difficulties in discovering where people are getting information and in coordinating efforts to focus on refining the more important information sources. There are number of researchers with the skills and knowledge to address these issues, but who may struggle to gather or process social data. The digital observatory facilitates data collection, access, and analysis.
Our initial release includes several datasets, code used to collect the data, and some simple analysis examples. Details are provided on the project page as well as our public Github repository. We will continue adding data, code, analysis, documentation, and more. We also welcome collaborators, pull-requests, and other contributions to the project.
What’s the goal for this project?
Our hope is that the public datasets and freely licensed tools, techniques, and knowledge created through the digital observatory will allow researchers, practitioners, and public health officials to more efficiently gather, analyze, understand, and act to improve these crucial sources of information during crises. Ultimately this will support ongoing responses to COVID-19 and contribute to future preparedness to respond to crisis events through analyses conducted after the fact.
How do I get access to the digital observatory?
The digital observatory data, code, and other resources will exist in a few locations, all linked from the project homepage. The data we collect, parse, and publish lives at covid19.communitydata.org/datasets. The code to collect, parse, and output those datasets lives in our Github repository, which also includes some scripts for getting started with analysis. We will integrate additional data and data collection resources from Pushshift and adjacent projects as we go. For more information, please check out the project page.
Stay up to date!
To receive updates on the digital observatory, please subscribe to our low traffic announcement mailing list. You will be the first to know about new datasets and other resources (and we won’t use or distribute addresses for any other reason).