CIVIC Structured Context Program

Have you read Data Feminism? Are you excited to put the ideas into action and connect with like-minded others?

We're building data systems that aspire to embody intersectional data feminist principles by creating the methodology and technology for collecting, documenting, and sharing contextual information about public datasets.

As a non-profit that makes data accessible for public use, context is a key part of our process and technology infrastructure and a key area where we see the potential for change. One aim is to highlight this contextual information wherever data communication is happening — whether that's an interactive data visualization, or a landing page for a dataset.

Join us for a 6-week experimental online applied learning and participatory experience!

Who it's for

Librarians, data scientists, designers, academics, journalists, educators, writers, programmers, artists, advocates, community organizers, program coordinators, information designers, metadata enthusiasts, poets & more.

This program is for anyone who has read Data Feminism and is interested in putting those ideas into action and connecting with like-minded others! We're designing this experience to accommodate a wide range of experiences, skill sets, and expertise, and will have different types of opportunities for different levels of experience. The problems and ideas we'll be investigating affect everyone around the world, and we welcome people from all countries.

Who we are

Civic Software Foundation is a non-profit working towards technology that can equitably serve the public. We facilitate interdisciplinary teams towards projects that make a difference. Through our Hack Oregon program we've worked with over 1000 contributors on 30 different projects over 6 years. Through that work, we identified common challenges, leading to the development of the CIVIC Platform, open-source technology, collaboration frameworks, and managed cloud environment to help make public information public knowledge.

The opportunity and the challenge

Quoting extensively from Chapter 6 of Data Feminism:

Governments and data providers have not invested as much time and resources in providing context to end users as they have in providing data.

And datasheets for data sets are great, but can we expect individual people and small teams to conduct an in-depth background research project while on a deadline and a budget?

So, intermediaries who clean and contextualize the data for public use have potential (and have fewer conflicts of interest), but there would have to be a funding mechanism, significant capacity building, and professional norms-setting that would need to take place to do this at scale.

We're ready to take on this challenge, and building towards taking it on at scale, with a structured context approach. We'll be putting this approach into practice for all future projects that make data available through our platform and continuously working to improve it.

Structured context is machine-readable metadata at the dataset level about what it is, where it came from, and who collected it, how it was collected, and why it was collected. Our structured context documentation aims to provide some useful machine-readable elements for dataset users, while not flattening details that are important to understanding. We aim to identify bias, ethical concerns, omissions, and other key considerations for use.

Here are some of the components that are part of this process:

• A context gathering research process guided by a common set of questions

• A contextual metadata schema that includes a mix of controlled vocabularies for useful classifications and more freeform information

• APIs and visual interfaces to make structured context information useable and accessible where data communication is happening

We've surveyed the current approaches to documenting dataset context, and we've found that highly structured approaches miss key areas of questioning, and that more comprehensive approaches could benefit from additional structure to facilitate re-use.

Learn more about our previous work in this area:

• Our questionnaire

• Example responses on Civic Platform

Our approach

Over the next 6-9 months, structured dataset context is core to our organizational focus and platform roadmap. We're creating an open framework for documenting contextual metadata for publicly accessible datasets, along with an accompanying team process and user interface for datasets hosted on CIVIC Platform.

This initial 6-week experience will be an immersive exploration of current approaches and future possibilities for structured context documentation. We will be using the learning from these 6 weeks to inform the rest of our 6-9 month build cycle. This 6-week experience is self-contained, but there will be opportunities for those who are interested to take on project roles in the build cycle.

As an open framework, other organizations can use this framework to document the context of their dataset and make it available in a standard format. We'll be using this process for all datasets that we onboard. Our aim is to highlight this contextual information wherever data communication is happening — whether that's an interactive data visualization, or a landing page for a dataset.

Timeline & time expectations

The initial learning & workshop series will run for approximately 6 weeks beginning the week of June 22.

We will have synchronous and asynchronous opportunities to participate. There will be a synchronous opportunity at Friday 12PM EST, the same time that the Data Feminism book club has been meeting. If you've been participating. You should expect a 2-5 hour / week time commitment, but with the opportunity for more if you have the appetite!

We will be conducting a 30-minute interview with interested participants to inform the workshop design.

There is no monetary cost to participate.