If you attended the Pittsburgh Technology Council’s recent conference on big data, you most likely were corrected at least once on pronunciation: not Day’tuh (long A) – but Da-ta (short A). Otherwise the joke of the event name wouldn’t work – “I Love It When You Call Me Big Data” – taken from the rap song of a similar name.
Pronunciation aside, there was a lot of big data trends and best practices to digest at the all day summit last week. The summit encouraged attendees to explore strategies and technologies surrounding real time data processing, data protection and privacy, industry regulations and compliance, and data storage.
Founded in 1983, the PTC is a non-profit trade association with more than 1,300 member organizations and more than 1,200 individual members. According to Matthew Holjes, vice president of business development, the council received offers from 45 potential speakers when the word went out that it was planning a conference– far exceeding the time allotted for a one-day event.
Ultimately, the PTC came up with the idea of “Tech Talks” – based on the format of TED Talks – with multiple 18-minute tracks and a plentiful array of presentations on topics ranging from "Oil and Gas Exploration's Big Questions for Big Data” to “Optimization, Data, and Scheduling Major League Baseball.” In other words, a little big data for everyone.
Most of the companies represented at the conference were Pittsburgh-based, although two of the morning speakers came in for the event – Rachel DiCola, senior director of research and consumer insights with Gamut, a New York City-based company formed by combining Cox Digital Solutions and Cox Cross Media, and Jason O’Connor, vice president of analysis and mission solutions with Lockheed Martin in Washington, DC.
The summit was co-sponsored by Rhiza, a marketing analytics firm. And what would any conference on big data be without some numbers? There were more than 200 attendees, 27 speakers, presenters and panelists, three hours of breakout sessions and six sessions for each of the three tracks with eighteen total presentations
Before DiCola’s panel, a crew curiously began passing out various bags of chips. DiCola explained that her presentation was based on research prepared for one of their clients and addressed the question “Are snack preferences regional?”
Maps and graphs zipped by at a breakneck pace along with DiCola’s accompanying details about each slide:
- Everyone in the country seems to like candy and potato chips.
- Cookies are favored in the southeast long with Texas and the mid-Atlantic.
- The northeast region, especially Pennsylvania, has a taste for pretzels.
- New Yorkers, on the whole, eschewed nearly all the snack groups, ultimately preferring – of course – bagels.
Income and age, personality and politically based munchies were all rattled off one after the other. The moral of the story for all this data: “You can have a thousand data providers. But if you don’t have a tool that helps you bring that together you can get buried very quickly.”
Keynote with Your Salad?
Jerome Pesenti, vice president, core technology at Watson Group, IBM’s Pittsburgh offices, gave the keynote. He opened with some clips of Watson, the supercomputer, stomping previous champs Brad Rutter and Ken Jennings on Jeopardy to the bemusement of host Alex Trebek.
The emphasis for Pesenti was on what Watson does and could do with big data – particularly in the health care field, by showing a video presentation of the possibilities for a hypothetical oncologist and his patient. Scenario: a doctor prepares to see a cancer patient by utilizing the IBM Watson Oncology Diagnosis and Treatment Advisor for relevant research through medical journals, up to and including suggestions for treatments and clinical trials. Pesanti emphasized that the supercomputer was an advisory tool, not a replacement, for physicians.
He also spoke about applications for deep learning, image recognition and speech recognition and translation. He included including the obligatory anecdote about Frederick Jelinek, a pioneer in speech recognition, who said, “Every time I fire a linguist, the performance of the speech recognizer goes up.”
Pass the Popcorn
Jeff Tabor, senior director for product management and marketing, Avere Systems, a Pittsburgh-based enterprise storage and data management company startup, neglected to provide the popcorn for his talk emphasizing Hollywood’s role in big data and cloud storage.
“Making modern movies is a pretty high-tech operation today,” said the man whose company helped to bring both the Hobbit and Zero Dark Thirty to the proverbial big screen.
Avere’s role is in the post-production phase – hosting the storage needed to perform the rendering and compositing of computer-generated special effects. In the case of Zero Dark Thirty, for example, a scene involving a sandstorm was filmed with neither storm nor sand. The creation of the digital effects for this scene required approximately ten thousand computers in a “render farm,” what Tabor referred to as a “supercomputer array of computers” with a disc storage space of multiple terabytes.
And it doesn’t take a wizard like Gandalf to see that the need for more storage is growing exponentially to keep up with the constant advancement of movie-making special effects. Envisioning as many as 20,000 to 40,000 computers required by 2016, “it’s just not sustainable to build that all out in their data centers so they’re turning to the cloud to achieve this type of scale.”
Of course, Tabor’s spoken presentation was a bit overshadowed when they played the trailer from “The Hobbit: The Desolation of Smaug.”
It’s hard to reclaim a stage when you’re following a dragon.