Datavisualization.ch has grown to be one of the most relevant and popular websites in Europe dedicated to developments in the field of data visualization. Following trends and reviewing technologies, it is a useful information source both for those interested in entering the field of data journalism and for those data journalists who want to be up to speed with developments in the field of data visualization.
DataDrivenJournalism.net interviewed Benjamin Wiederkehr, founding partner of datavisualization.ch and of the design studio Interactive Things, to pick his brain about everything a data journalist should know about tools, trends, dos and don’ts in data visualization.
What has changed in data visualization since you started the website in 2008?
Generally speaking, the greatest change concerns the very popularity of it, the breadth of the topic and the number of people involved. There is a far greater number of businesses and industries that use data visualization as a way to analyze, evaluate and communicate their information. When we started, the publication frequency of new material was quite low, while, nowadays, new data visualization projects are published every other day.
Data journalism, in particular, has seen a very steep curve of increased attention and output. This is particularly true in the case of interactive data visualization, I should say, such as the work done by The Guardian or by The New York Times, which are fully engaging interactive media as a way to support their articles, as well as to attract increased interest from the general public.
What are the hottest trends in data visualization today?
Technologically speaking, this is a very interesting time for data visualization, because the tools that allow people to create visualizations, communicate information and start telling stories with data have changed dramatically over the past years. There are a few tools available that allow you to publish data in a very engaging way, while also doing justice to the data itself. I’m thinking about tools like Tableau and Tableau Public, but also the inclusion of charting possibilities within Google Docs - a development that has helped a lot of people to get started without a lot of programming knowledge.
There are also programming languages, or libraries, for creating visualizations through processing or through JavaScript, that have changed dramatically. D3, for instance, is very popular at the moment and seems to be the de-facto way to publish visualization on the web. It has completely taken over the work that was done with Flash in the past.
There are also new trends from a visualization perspective, and by that I refer to the work done with real-time data, a data feed that is ever changing as a source for data visualization, like a Twitter feed, for example.
Another trend involves letting the users create visualizations and share their customized view on a dataset with their peers or friends. The involvement, the dialogue between the data and the creator of the visualization, on the one hand, and the users themselves, on the other hand, is gaining popularity.
What are your favorite data visualization tools for journalists at the moment?
My absolute favorite is the D3 library, which I mentioned earlier. It is a flexible tool and the community around it is very active, which leads me to believe it is here to stay. But there are other tools that build on D3, like Datawrapper. It is actually useful for journalists who want to publish visualizations based on data they already have, making it easy for them to create D3 visualizations.
Tableau is another, rather sophisticated way to publish visualizations, especially if you put together a collection of visualizations. The results can be very nice. We rely heavily on Tableau in our studio as a means to help us understand the data at the very early stages of our work, to explore it before we move into more customized productions.
We sometimes use Google Refine, now called Open Refine. It’s not a visualization tool, but often visualization work involves refining and understanding the data. Open Refine helps us to get a sense of the texture and structure of the datasets we are working with.
IBM created a visualization tool called Many Eyes. It is a web application that allows you to upload a dataset and to explore it with many different pre-built visualization techniques. It goes a bit beyond the possibilities given by Tableau, but it is also a little bit more restricted when it comes to combining different charts into full-fledged dashboards. But still, it is definitely interesting.
Quadrigram is another recent, interesting tool. It is the new version of a tool that has been created a few years back called Impure. It is an application built in Flash, and maybe for that reason it was not received as enthusiastically as it should have been. But the tool itself is very interesting. It is a sort of prototyping tool, or a production tool, for working with data. In a broader sense, it allows you to connect data sources of different kinds, enabling you to transform the data and to visualize it in the same application. You can do it in a very sophisticated way without programming code. It is a very interesting tool. I highly recommend giving it a try.
Which tools are more suitable for beginners, and which for veterans?
The best ones for beginners are Tableau Public and Many Eyes. These are straightforward tools to start playing with. And Datawrapper, definitely - it guides you through the process and that is something I find interesting.
Another one is MapBox - a mapping application. Mapping applications are very popular. With a very simple system and a simple user interface, MapBox allows users to upload data, to map it in many different ways and to create maps that tell the story. It is a very accessible tool for beginners.
As far as the more advanced tools are concerned, D3 is currently the framework that allows you to do very customized and sophisticated things for the web. There is also Processing, which is still commonly used, though not so much for the web. But if we work on projections or on touch screen interfaces that are large scale, we may use Processing because it is built with Java and therefore has a lot of advantages over regular web technology.
And then there’s also Gephi, which I would also recommend for advanced users, specifically for networked visualizations - social networks, networks within companies or between companies.
What distinguishes good data visualization from bad data visualization in journalism?
Good visualization really supports the story. A bad visualization lives on its own, outside the narration and the context that it originated from. If the story is pre-written and the visualization is sort of separate, it is not put to good use.
I would like to see deeply interconnected journalism, where the written word, the interactive visualizations, images and other multimedia materials are put together to form a narrative that really creates a piece of journalism. I would love to see more of this interplay between different kinds of media. If we manage to make a guided interactive experience for users, we will be able to tell very interesting and very engaging stories.
Do you think the popularity of data visualization will ensure a more prominent role for graphic designers in the newsroom?
I would expect to see more interdisciplinary teams that include journalists, designers and developers, like a hybrid of two or three disciplines. There are tasks in the workflow of data journalism that require bringing together different types of expertise to create something of high quality. I hope, therefore, to see a closer collaboration between journalists, designers and developers.
In the traditional approach, a journalist does the research and writes the story, to which a few images are later added. If you want to work on interesting visualizations that are fully embedded in the story, you need to talk early on with the visualization designers or developers, discuss with them the visualization method, how to represent the data and how to combine it with the textual parts of the article.
We are still far from having a fully satisfying method for reaching an optimal interaction. A lot of people are currently exploring the means to get there and how to best engage and attract users towards stories that are told through interactive visualization.
Who are trendsetters we should monitor in the future?
The usual suspects immediately come to mind: The New York Times constantly comes up with good examples, which amaze me, especially when I get a peek at the work done behind the scenes and their setup. They have built a fascinating team of very competent people that seem to collaborate perfectly.
The Guardian also does an incredible job quickly producing visualizations that support their stories. I also recommend looking at the Malofiej Awards that reward the best people in information design. Interactive visualization for journalistic purposes is always a part of it. So looking at what Malofiej rewards is a good starting point for finding the next best thing. It is also a good place to watch last year’s recap of data visualization production.
I also like to check conferences, like the see conference that will take place in Germany in April, or the Eyeo festival in Minneapolis, among others. I like to attend, when I can, and hear the speakers talk about how they work. Visualization projects are published and are easy to find. But learning about the process that led to these results - this is what I’m most passionate about.
_Moran Barkai is the editorial assistant for DataDrivenJournalism.net. She has worked for publications such as Haaretz, the Huffington Post, Time Out, Your Middle East, and the European Journalism Centre's Web-magazine._
_This post originally appeared on DataDrivenJournalism.net and is posted on IJNet with permission under a Creative Commons Attribution-NonCommercial license. Created by the European Journalism Centre, this data journalism initiative is aimed at enabling more journalists around the world to use data to improve reportage._
Image CC-licensed on Flickr via mkandlez.