Fresh opportunities in processing mega-sized data sets

By Kim Tae-gyu

Big data is evolving fast. After replacing cloud computing of late to become the most frequently touted buzzword in the information-technology world, it generates a new possibility in dealing with mega-sized data sets.

Altair Korea Representative Director Moon Song-soo called the trend big data 2.0 _ the second-generation processing of large-sized data for improved efficiency and actual usage.

``Up until now, companies have tried to just come up with programs of working on big data because the demand for such services shot up over the past several years,’’ Moon said in a recent interview.

``Beautiful graphs showcasing the data are important. But more significant is how to provide a set of tools such as what-if analyses and price predictions to help decision makers. That is big data 2.0.’’

During the big data 1.0 period, Moon pointed out that the most significant topic people cared about was the capacity of processing data.

With the advent of the big data 2.0 era, however, he expected that an increasing number of users needing big data tools would realize how crucial the optimization or what-if analyses are, in the near future.

``We have struggled to manage huge data sets, which prompted experts to coin the terminology of big data even though it is uncertain who created the word,’’ the Altair Korea founder said.

``Yet, the genuine purpose of managing such mega-sized data sets is to process and analyze them. I think that optimization and what-if would be on the lips of people beginning late this year or early next year.’’

Analysis vs. intuition

A manager in a family restaurant in Jeonju, around 240 kilometers south of Seoul, had a knack in predicting what people would order in different seasons and weathers amid fast-changing and elusive consumer tastes.

He sold approximately 80 percent of food materials he purchased, way higher than the average, which is in the vicinity of 50 percent. The remainders are mostly wasted to weigh on financial status of restaurant operators.

A few years ago, the head office of the family restaurant chain closely studied the uncanny ability of the extraordinary manager based on interviews, statistical tools and other means available.

The intensive examinations enabled the birth of a defined model, which eventually jacked up the usage rate of purchased foods to 72 percent from some 50 percent without the model.

``When the stories spread, people tended to conclude that human intuition is better than analysis of computers because the latter could not match the former’s norm of 80 percent,’’ Moon said.

``However, that might not be the case. The expert at issue might be good in his own restaurant and it remains to be seen whether or not his magical intuition would work in other places.’’

By contrast, he noted that the analysis based on big data substantially cranked up the accuracy of restaurant managers’ prediction on their sales in a reliable fashion in any of the chain’s other branches.

In particular, technologies of grappling with big data sets have dramatically got better during the past few years to provide leverages to machines in their competition with humans.

``I would recommend decision makers to rely on data rather than human ideas. By now, big data and its analysis would offer better chances compared to ideas of folks thanks in no small part to big data processing,’’ he said.

Many seem to buy Moon’s suggestions because not only corporate clients but also politicians, governments and civic groups continue to knock on the doors of big data companies like Altair Korea.

In particular, political groups have briskly sought after analyzing big data to understand bona-fide public opinions through delving into vast data sets generated through the Internet and mobile.

Migration to big data _ life log

Large and complex data sets take center stages in so many sectors, which can’t be worked on with on-hand database management tools and this leads to the enhanced roles of big data-tailored software and sufficient computing capacity.

The trend to larger data sets is likely to accelerate since the data size continues to grow thanks in no small part to ubiquitous usages of mobile devices and location-based information.

The number of mobile phone subscribers amounts to 6 billion out of the world’s total population of 7 billion and many of them carry smartphones and other always-on Internet-enabled devices such as tablets.

Against this backdrop, mobility-specific services and applications like Facebook and Twitter keep gaining popularity to generate huge amount of data.

Looking forward, Moon said that big data can be applied to each person, who wants to build a life-long data like life-loggers who captured all his physiological information with a wearable camera.

``You have lots of video clips of your kids and the amount would keep increasing. All the data can be a personal life log, which is kind of big data when the data of folks accumulate,’’ he said.

``Then new opportunities emerge. For example, we would be able to analyze when Koreans typically come down with a specific disease. There would be some privacy issue to be addressed, though.’’

Engineer-turned-CEO

After studying engineering at Hongik University, Moon started his professional career at Kia Motors in 1996 where he worked on a simulated crash test of new vehicles of the Seoul-based automaker.

Then, he co-founded Altair Korea in 2001, a subsidiary here of the U.S.-based data engineering firm with a knack in engineering consulting, computer-aided engineering and business intelligence consulting. Most recently, it has made a dent in a big data market with its cutting-edge software and solutions.

During the first decade of the new millennium, Moon has catapulted Altair Korea to become one of the fastest-growing affiliate of the global engineering giant, which has a presence in 16 countries.

Ever since its debut, the outfit chalked up an average of some 20 percent growth in sales while it somewhat languished in the late 2000s in the midst of the global credit crunch and financial crisis.

In the early 2010s, however, the Seongnam, Gyeonggi Province-headquartered firm returned to a double-digit growth per annum. It aims to reach the 10 billion won mark in turnover next year.

Included in its customers are such top domestic companies as Samsung Electronics, Hyundai Motor, LG Electronics, Hyundai Heavy Industries and GS Caltex, which use the firm’s enterprise solutions.