CSE Speaker Series – Max Planck

On Friday, February 24, in Cramer 221 at 11:00, Max Planck of ICASA will give a talk on DAVE.

“A Problem With Architectures – How Big Data made “easy” makes analytic development hard” – This talk focuses on a specific challenge encountered during the building an analytic development system called the Data Analysis and Visualization Environment (DAVE). The purpose of DAVE is to allow non-programmers to compose data flow based analytic processes against many disparate Big Data sources. The problem is that while any single step in an analytic process can be efficiently mapped to Big Data programming paradigms (i.e Map/Reduce, Bulk-synchronous Parallel, Spark, Kafka, Storm, Columnar/Graph Databases) it appears to be difficult to do the IO between steps in such a way that efficiency is preserved. This talk tries to present the problem, and some initial ideas about how to deal with it.