“Our goal is to automatically remove large inefficiencies in [big data] systems, significantly improving their efficiency and making them scale to larger data sets,” says Harry Xu, associate professor of computer science.

Harry Xu, associate professor of computer science at UCI, and UCLA colleagues have been awarded $4.9 million from the Office of Naval Research to support their work on reducing software inefficiencies. The four-year grant will fund the development of software customization techniques that can greatly reduce the redundancies and waste in modern, object-oriented big data systems, said Xu, who’s one of the pioneers in software bloat analysis. The techniques will be applied to widely used Java software — in particular, to large-scale data analytical systems such as Hadoop, Spark and AsterixDB. According to Xu, these systems have become increasingly important in modern big data computing because they form the backbone of next-generation cloud computing and internet services. “Our goal is to automatically remove large inefficiencies in these systems, significantly improving their efficiency and making them scale to larger data sets,” he said.