Subscribe / Unsubscribe Enewsletters | Login | Register

Pencil Banner

Startup offers a big-data storyteller

Joab Jackson | Oct. 20, 2014
Amassing a pile of big data is not sufficient to spur insight, even if you have clever ways of analyzing that data. The savvy data scientist must also find a way to make all this information tell a story so it can be understandable to others.

To use the service, the user uploads one or more sources of data — such as a CSV file, spreadsheet or relational database — or provides an API link to a live data source. The service then offers a number of ways to combine and visualize the data or do some basic mathematical operations, such as finding averages.

In order to ease the ingestion of data from multiple sources, the company built what it calls a data inference and profiling engine, which can make many basic assumptions about how a new set of data should best be formatted. The service also builds a set of metadata, covering aspects such as time, or location.

The service relies on a proprietary bit of code the company calls a harmonization engine, which looks at the metadata for common points across different data sets, zip codes or time periods, for instance. From this work, it produces recommendations that, given a data set, offers suggestions about what other data sets could be integrated. The resulting graphics, or set of visualizations, can be bookmarked for others to see.

While the original service was aimed for business analysts, the new Collaborative StoryBoards is marketed towards management and senior executives, allowing them to combine multiple contributions into a single presentation.

With storyboarding, a team leader can start a project and multiple contributors can add their own graphs and visualizations. Each member of a team can also comment on the individual stories. When all the data is collected, the work can be posted, used as an extended live dashboard of sorts and shared with others.

Pricing is based on an annual subscription model, including the amount of data being processed and the number of storyboard authors. A typical enterprise deployment may start at about US$50,000 per year

The idea of using stories to interpret large amounts of data has been a theme at the Strata + Hadoop World conference. Mapping startup CartoDB offers a way to visualize geospatially oriented data by overlaying that data on maps. Storytelling has been an essential part of this service, explained CartoDB senior scientist Andrew Hill in a presentation.

"Really, we're trying to help people gain insights and tell stories about their data," Hill said. Storytelling is essential "if you want to make a point, or you want to convince other people about the importance of what you're doing."

Venture capital firms such as Andreessen Horowitz, DAG Ventures, Google Ventures, Khosla Ventures and Kleiner Perkins Caufield & Byers have all invested in ClearStory Data.

 

Previous Page  1  2 

Sign up for Computerworld eNewsletters.