Data Discovery

Discovering relevant datasets can be a challenge, often as a result of not knowing what data is needed because we don’t have any idea what might exist, and not knowing what to do with data we have because we don’t know who might want it (and what they might do with it).

The solution, as is regularly the case, is to convene a group of people to explore it together. The good thing is that this doesn’t have to be everyone who might have or need data (after all, we don’t yet know who they are): it’s enough to gather a few people with some knowledge of the theme, issue or opportunity; between what they know and who they know it’s generally possible to build quite an extensive picture of the data landscape.

This post outlines an approach to data discovery that we’ve used with various clients. We’ll be adding some materials to our Github shortly (along with materials for other tools we use), but for now here’s a description of how we generally go about this kind of work.

The Components

The important thing is to build data discovery work on a base of people, organisations and interfaces. It’s important to have both people and organisations because it stops the workshop getting bogged down with uncertainty around who the right people are in an organisation, or exactly which organisation somebody works for.

Interfaces can be anything, from websites to back-office systems to sensors to notebooks (in a recent workshop with librarians we discovered a treasure trove of qualitative data that had been written down after conversations with library users).

Core Activities

On its simplest level a data discovery workshop is all about coming up with ideas for those people, organisation and interfaces, writing them on cards and spreading them out on a large table. There are many ways to configure the cards, for example putting the most important ones in the centre, or describing a few different themes by clustering groups of cards together. We’ll share some examples of this approach in use in a later post.

Once you have a table full of cards you can start to think about the data that connects them. What data do they produce or consume? What data might you be able to get from them? What data do they need?

Asking those questions, writing the answers down and revisiting your map will lead to new questions. You’ll want to add in more people, organisations and interfaces. Or take some out.

This iterative process is how it works. Ideas lead to conversations, which lead to a better map, which leads to more ideas and conversations. Allowing the space for this to happen is how you get the most from the people in the room and how even a small group can build a sophisticated map.

Enhancements

You’ll get a lot from the workshop described above, but depending on your audience you might find that some additional activities help extend the reach of the workshop. For example, if you need to give people some additional information to inspire them it’s worth kicking the day off with a series of lightning talks from people with deep knowledge of the themes.

Or you might know that the people you’re working with have some good existing knowledge of the data the have or need so some round table discussions at the beginning can help uncover that first.

However you approach it, the most important thing is to think about the participants and what they come with or what they might need to really dig deep. We’ll share more examples and ideas around this in future posts.

Permanent Installations

The approach here can also be used as part of a permanent installation, in an office for example. By using materials that attach to a wall you can make data discovery an ongoing process. This might be cards, pins and string or similar.

Interchangeable Materials

We’re working on making the materials from these workshops interchangeable with materials we use in different types of workshops. Over time we’ll be building this into a kit that you can read about on this Medium Publication and download from our Github.

The Data Place helps people and places thrive through better data discovery, publishing and use. We’re a social enterprise, bringing together infrastructure, data skills, human-centred design and open source development. Find out more: https://thedata.place/.