By John Cartan, Oracle Applications User Experience
Big data has become an important force in the enterprise, one of special interest to Oracle. But in order to make data usable you must first understand it – in short, you need some way to see it.
In the last few years, the Oracle Applications User Experience (OAUX) team has been exploring new ways to visualize data, especially big data. As scales increase, everything becomes harder: gathering the data, cleaning it, organizing it, even just drawing it.
The amount and variety of data to gather has exploded, due in part to the rise of the API economy. Companies have discovered it’s profitable to make their private data public, and they can provide APIs, or application programming interfaces, to make it easier for other people to retrieve that data.
This project aims to visualize nine years worth of “Likes” to author John Cartan’s Facebook posts. Read more about the project on The AppsLab blog.
Even if they are not participating in the API economy as providers, Oracle customers will increasingly need to access semi-structured public data from social media and other sources to enhance their own internal reporting. So the AppsLab, the OAUX emerging technologies team, has focused on projects such as Fun With FaceBook to explore how to access and process data from public APIs.
Before data can be visualized, it must be inspected and cleaned. This requires "pre-visualization," seeing and shaping raw data to spot anomalies or just to figure out what you've got. For designers trying to turn that data into dashboards or maps or animations, there is a more subtle problem: the more data you have to work with, the harder it is to experiment and build prototypes. The next generation of designers will need new tools, a challenge discussed in Better Ways to Play and Try.
When designing any kind of interface, the OAUX team’s mantra is "glance, scan, commit." For big data this means finding ways of seeing both the forest and the trees: overall structure at a glance with the ability to effortlessly drill into details when needed.
Another recent post, Who Likes Me?, describes a project to research large-form, high-resolution information displays for possible use in video walls and control centers, for example. Pushing the envelope exposed current limitations that were both technical and ergonomic.
Even as speed, memory, and cost continue to improve, visualizations will still need to satisfy human limitations. The differential between tree and forest must always fit comfortably so that no magnifying glass is needed for details and no stepladder is needed to see the big picture.