Alessandro Margara
Lecturer Alessandro Margara
Title Managing streams of data: lessons learned and open challenges
Date Monday 09/05/2016, 3.30 – 4.30 p.m.
Location S2|02, Room C120
Rundeturmstr. 10, 64283 Darmstadt
Abstract
Many modern software applications involve processing, analyzing, and reacting to potentially large volumes of streaming data. Examples of such
applications include monitoring systems, decision support systems, financial analysis tools and traffic control systems. Designing and implementing these
applications is difficult. Indeed, the streaming nature of the data demands for efficient algorithms, techniques and infrastructures to analyze the incoming information on the fly and extract relevant knowledge from it.
Bio
My research activity focuses on designing programming abstractions and tools to help the developers in dealing with streaming data. During this talk, I
will present my experience in this area. I will first focus on the design of a language explicitly conceived to identify situations of interest from large streams of low level data. Next, I will discuss the implementation of efficient processing algorithms for streaming data that can exploit modern many-core architectures, such as multi-core CPUs and programmable GPUs, to reduce the processing latency and increase the overall throughput of the system. Finally, I will present the open issues and challenges in the field of stream processing and draw a roadmap for future research in the area.