Inside the Newsroom

How would you change marijuana laws? 

That is a question the Center for Investigative Reporting and KQED posed to readers following the Obama administration’s letter to U.S. attorneys reminding them that the cultivation and distribution of marijuana is illegal under federal law. The answers we received were thoughtful and thought-provoking, but we struggled with the best way to present them. Then our partners at the Public Insight Network announced they’d created Skyline, an interactive visualization of responses from sources in the network, and were looking for a way to unveil it. I jumped at the opportunity.

To properly explain what Skyline is and why it is so cool, I asked its creator, Barrett Fox, to share a few words with you. Fox is a data visualization designer at American Public Media working on the Public Insight Network. He has worked throughout his career at the odd but thrilling intersection of video games, animation and information visualization:

To view the results of our query on Skyline, click here. For a handy instructional video on how to use Skyline, click here. Skyline can be viewed on a PC or Mac. First-time viewers will be prompted to download the Unity 3-D browser plug-in.

Every day, journalists at California Watch and the Center for Investigative Reporting and at newsrooms around the country ask the more than 130,000 sources in the Public Insight Network to share their knowledge and insight on any number of topics of public interest. The result is a formidable amount of largely unstructured data, and this huge info cloud quickly becomes overwhelming.

The goal is to synthesize this information quickly – to make sense of it in a way that adds context, depth and relevance. So, how to do that?

Sure, we can search it, but you'd have to scroll down web pages that are miles long to get a clear sense of it all. It’s easy to lose track of the most important themes and to get lost in the details. We believe there’s value in being able to quickly see the broad strokes of these conversations, to paint a portrait of what we’re learning as we learn it.

This is where the rapidly evolving and expanding field of data visualization or, as I like to call it, information visualization, comes in. Not many years ago, data visualization was a field of charts, graphs and, if you were lucky, some diagrams. But today, we can quickly transform mountains of data into the liquid medium of interactive computer graphics, presenting us a new palette of possibilities.

Skyline is an interactive visualization of responses from sources in the Public Insight Network. It’s intended to quickly show how many people responded to a PIN query, how much they had to say and then, when you "drill down" into the Skyline, to allow you to read those responses in detail. The joint CIR-KQED query about marijuana laws is the first Skyline visualization to go live, and as you'll see, people had a lot to say on this topic. Mind you, it’s a prototype – and we want your feedback to help us take it from a cool concept to the super-useful information explorer we want it to become.

Skyline is built on the Unity 3-D game engine that provides a world of new possibilities for interactively and visually exploring information. Instead of data about how many monsters you've clobbered or gold coins you've collected, we're feeding this game engine data from the PIN queries. And then we can design how that information gets displayed with all the tools that video game designers have at their disposal.

By creating a 3-D icon for each source, we can quickly see the size of the crowd of people who wanted to talk about this subject. And by stacking their actual responses above them, we can quickly see how much they all had to say. And we can group the source based on specific questions in our queries to reveal otherwise hidden qualities about the group of respondents. For this first example, we're grouping the responses by their ZIP codes, which in this case shows that we heard from people around the state.

We’re building Skyline to visualize any query in the PIN and create groups based on any questions in those queries. And what you're seeing here is our first experiment. But what's exciting is that now that we have these sources and their responses inside this interactive 3-D medium, we'll shortly be able to explore this data in many other interesting ways. We can discover other insights on these topics by placing the sources on maps, timelines, even good old charts and graphs. But you'll always be able to drill down and read what's important – the knowledge and insights of citizens who are helping make the news more credible, relevant and transparent.