Watson, a computer program named for IBM’s founder, defeated the two best human players on “Jeopardy!” one of the most challenging game shows of all time, in 2011, marking a significant milestone in the advancement of big data analytics.
Since then, big data has spread into many academic disciplines and industry practices. Earlier this year, three alumni of the University of Houston's Bauer College of Business MBA program won awards for papers on big data in analytics. Facebook began analyzing social media activity to target consumers with advertising, and grocery stores started examining individual purchasing patterns to determine product placement.
“It’s good that we have access to lots of information, but it’s too much,” said Haluk Ogmen, professor of electrical and computer engineering and director of the Center for Neuro-engineering and Cognitive Science at the UH Cullen College of Engineering. “The idea is to use big data analytics to extract the useful information.”
Shivakumar “Shiv” Vaithyanathan, chief scientist for big data analytics at IBM Research, will talk about ways to make that happen during a presentation on the UH campus at 4 p.m. Oct. 24 in the Classroom and Business Building, room 106.
Vaithyanathan, an industrial engineering alumnus of the UH Cullen College of Engineering, also manages the Machine Learning Systems Group at IBM Research. His work is at the intersection of natural language processing, machine learning and databases.
He was named an IBM Fellow earlier this year.
Only 257 employees have earned IBM’s highest honor since Thomas J. Watson Jr. founded the IBM Fellows program in 1962. The company employed more than 400,000 people worldwide in 2013, according to the IBM website. The Fellows have generated 7,700 patents collectively, and five Fellows have won Nobel Prizes.
In his talk, Vaithyanathan will briefly introduce several applications of big data analytics technology to tasks ranging from investment and equity research to social media lead generation.
He also will describe major analytic phases at the core of the applications. The phases include text analytics, semi-structured data processing - including joins, group-by and aggregation operations - and statistical and predictive modeling.
“At IBM, we are building tools and technologies to support each of these analytic phases,” Vaithyanathan said. “In particular, we are building declarative languages for these phases.”
Vaithyanathan plans to devote the second part of his talk to SystemML, which expresses ML algorithms in a higher-level language and compiles and executes them in a MapReduce environment. He will end the talk with a discussion of speeds, feeds and comparisons.
Before Vaithyanathan joined IBM, he was a founding member of the Altavista Group at Digital. He has co-authored more than 40 papers for major conferences including ACL, EMNLP, SIGMOD, VLDB, ICML, NIPS and UAI. He was a keynote speaker at the 2011 German Database and the 2011 ACM SIGIR Industrial Track Conferences. He was also associate editor for the Journal of Statistical Analysis and Data Mining from its inception until 2012.
“Big data analytics is a very contemporary area that is of interest to many people,” said Gino Lim, professor and chair of the Department of Industrial Engineering at the UH Cullen College of Engineering. “It’s in every discipline because you have so much data with the Internet, so now the limitation is how fast you can process such a large amount of data to extract useful information.”
Who: Shivakumar “Shiv” Vaithyanathan, IBM Fellow, chief scientist for big data analytics and manager of the Machine Learning Systems Group at IBM Research
What: “Declarative Big Data Analytics: Applications and Tools.” Co-sponsored by the Cullen College’s Center for Neuro-engineering and Cognitive Science and the Department of Industrial Engineering
When: 4 p.m., October 24
Where: Room 106, Classroom and Business Building, UH Main Campus