BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VTIMEZONE
TZID:Europe/Vienna
BEGIN:DAYLIGHT
DTSTART:20190331T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20181028T020000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260405T192556Z
UID:5c6a9fad16268811967880@ist.ac.at
DTSTART:20190318T100000
DTEND:20190318T110000
DESCRIPTION:Speaker: Ioannis Liagouris\nhosted by Krish Chatterjee\nAbstrac
 t: Distributed dataflow systems have become the de-facto systems for large
 -scale data processing. So far\, most efforts from academia and industry f
 ocus on improving the performance\, scalability\, and reliability of these
  systems. However\, as their complexity increases\, explainability emerges
  as a first-class concern.In this talk I will present my work on understan
 ding the behavior of distributed dataflow systems and applications. First\
 , I will describe a framework for explaining the semantics: Why and how do
 es a dataflow return certain results? To answer such questions\, the frame
 work leverages ideas from database provenance to provide output explanatio
 ns that are both sufficient and concise. Second\, I will focus on understa
 nding performance: Why is a dataflow execution slow and which are the bott
 lenecks in the pipeline? My work in this area generalizes existing approac
 hes on critical path analysis methods to dynamic and continuous computatio
 ns. I will conclude the talk with a discussion on the challenges of explai
 ning emerging AI applications and an overview of my future research agenda
 .
LOCATION:Mondi Seminar Room 2\, Central Building\, ISTA
ORGANIZER:tguggenb@ist.ac.at
SUMMARY:Ioannis Liagouris: Understanding distributed dataflow systems and a
 pplications
URL:https://talks-calendar.ista.ac.at/events/1817
END:VEVENT
END:VCALENDAR