The Power of Data

Estimated reading time: 7 minutes

"Students don’t all need to be crack programmers," says Daniel Lopresti, director of the Data X initiative and professor and chair of the Department of Computer Science and Engineering. "It probably behooves a lot of them to learn programming skills, but there are other kinds of skills that allow students to work with and manipulate data in powerful ways. [We also will develop] an awareness of data science and what the opportunities are, what the tools are, and then how to feed the system and interpret the results."

Through a large hiring initiative, Data X will expand its faculty working in computer science and related fields to make available to students of all majors broadly accessible courses that will achieve these goals.

The imperfect data of an imperfect world

In the early days of data analysis, businesses would collect structured data such as units and monthly sales and enter them into Microsoft Excel. Examining that data took time—an analyst would run functions, consider possibilities and try to recognize patterns.

"If you had perfect data in your spreadsheet," says Lopresti, "you could make some conclusions."

Today, "big" unstructured data is a major factor.

"We’re collecting megabytes, gigabytes, terabytes of data in very short periods of time about everything from industrial processes to business experiences to your personal experience, your entertainment experience," says Lopresti. "We’ve gone from basically a very data-poor environment to a very data-rich environment."

Much of this data is too large and complex for traditional applications to process. IBM breaks this "big data" into four elements: volume, velocity, variety and veracity. Volume refers to the sheer quantity of data available; velocity to the speed at which data is collected, organized, analyzed and decisions are made. The variety of the data collected has exploded in recent years. According to an IBM infographic, for example, 400 million tweets are sent daily by approximately 200 million Twitter users. This is in addition to the more than 4 billion hours of video viewed on YouTube each month and the 30 billion pieces of content shared on Facebook. The final "v," veracity, has to do with how much the data on which decisions are based can be trusted. How reliable is the information?

Big data is noisy and incomplete, says Lopresti. "So you need computational techniques that can deal with the volume, the velocity, the variety and the noise and the real-world aspects of [big data]."

People—intelligent, thoughtful, well-trained people—are helping the world get better and better at developing these techniques.

Page 2 of 4

• < Previous Page
• Next Page >