DEBS Day 1

Hola!

The keynote, Rise of the Velocity Pipeline, by Roger Barga of Microsoft, talked about the use of event-based analytics in business. In many ways it wasn’t very surprising, talking about the increase in the number of connected devices and so on. Rolls Royce was a prominent example!

The day continued with the research sessions. I won’t summarise them here as details may be found at

http://www.orgs.ttu.edu/debs2013/index.php?goto=program

The paper that really struck me was Ariadne: Managing Fine-Grained Provenance on Data Streams by Boris Glavic, Kyumars Sheykh Esmail, Peter M. Fischer, and Nesime Tatbul. Their goal was for each tuple to identify the tuples that were used to produce it. Given the operator graph that describes a streaming application, they have some transformations that add and modify the operators to generate, propagate, and filter provenance. The paper contains a few optimisations (unchecked, provenance information is huge!) and they do some experiments. Neat stuff and I intend to look at it more.

A bit late describing the industry session, aren’t I? I have to admit that I phased out a bit. The papers dealt with experiences using event-based systems for tasks such as process monitoring and mobile network QoS. Industry likes centralisation–no surprise there!

On to Tuesday…

This entry was posted in Big Data, Data Analytics, David Evans, Research, Uncategorized. Bookmark the permalink.