Saturday, July 16, 2011

On pattern rewriting - DEBS 2011 presentation



Among other things I have done in DEBS 2011, I have also delivered one talk to present the paper on patterns rewriting, the paper, like all DEBS 2011 papers can be obtained from the ACM Digital Library
The paper is co-authored by Ella Rabinovich (the primary author), Avi Gal and myself.
The presentation starts by classifying optimization tools in use for optimizing the performance of event processing applications, the classification is into blackbox optimization where the actual implementation is taken as a blackbox, and whitebox optimization,      The paper deals with one of the types of whitebox optimization - pattern rewriting, which mean rewrite a pattern into a collection of other patterns that yield equivalent results: same output to the same input. 

 While rewriting exist in different areas such as rules or SQL queries,  there is some inherent complexity in some of the event processing patterns such as SEQUENCE.    The difficulty stems from the temporal relationship imposed by sequence, which imposes restriction on assertion splitting, and from the fact that policies may not be transferred in the rewriting process nicely.   This was an exploratory research to see whether rewriting is a useful method for optimization, especially where complex sequences are involved.  The experimental results in simulation for sequence with 16 operands (a real application we have done in the past in the web commerce area),  shows that rewriting alone can improve the latency by more than tenfold and also the throughput, and no matter how the latency/throughput trade-off is weighted,  some rewriting will give better result than no rewriting in this case.   

The presentation can be viewed on slideshare.  Enjoy!

Summary of DEBS 2011



In this picture you can see me explaining the rules of  
the game for the "gong show", while Francois Bry, who was the gong master, is standing behind with the gong.   You can also see that the podium has the Jeopardy! setting, and indeed we had in one of the sessions live game of two of the participants against Watson. 
Paul Vincent, skilfully played the host, Scott Schneider and Michael Olson were the contestants, although the audience helped them Watson won this time also.


These pictures and many others were taken by Roland Stuhmer, and can be found on Flickr.


Besides the fun part we also had 33 papers, poster and demo session, DEBS challenges, tutorial day, PhD workshop and 4 keynote speakers


Overall,  the community that includes researchers and some people from industry is becoming stable. the terminology becomes more coherent (everybody are talking about patterns), and the quality becomes higher. 


There were participants from 23 countries from all the 5 continents including the first representative from Africa, a faculty member from Tanzania.    I think that still the largest community is in Europe.  


Organizing a conference is a lot of work, but is also a rewarding activity.   My next conference in early August - AAAI in SFO,  but there I am just a participant.     


More on DEBS 2011 -- later   







Tuesday, July 12, 2011

DEBS 2011 tutorial on non functional properties of event processing

This is the picture that opens the tutorial on non functional properties of event processing that has been delivered yesterday within the DEBS conference by Tali Yatzkar-Haham and myself.    It shows variety of various cakes, each of them can be entitled: "cheese cake".     Likewise, there is a substantial variance among event processing systems based on their non functional properties.  We surveyed a collection of topics on performance, scalability, availabilty, usability and security/privacy.   One of the conclusions is that the usability area lacks comprehansive research.     At the end of the day we went (with some other friends) to eat dinner in the cheese cake factory  in White Plains (I did not eat any cheese cake). 


The tutorial is now available on slideshare -  enjoy!



EPTS awards announced




DEBS'11 is quite a dense conference with a lot of activities,  the only free evening is the tutorials day evening, which we used to have a co-located EPTS meeting, whose primary focus has been to grant the EPTS awards.


I am copying the notification sent earlier to all EPTS members:




Today, at the EPTS session that took place in the evening of the DEBS'11 tutorial day in Yorktown Heights, NY.  EPTS announced its first series of EPTS awards.

Awards were given in two categories:    EPTS innovative application awards and EPTS innovative principles awards.  In each category two awards were granted.

The EPTS innovative application awards were granted  to:

  • Starview Technology for the work on Advanced Logistics and Planning Solutions - ALPSTM
  • Roberto Baldoni, Gregory Chockler and Giorgia Lodi for the work on CoMiFin

The EPTS innovative principles awards were granted to:
  • Hans-Arno Jacobsen and Mohammad Sadoghi for the work on BE-Tree: Boolean Expression-Tree
  • Elke Rundensteiner, Di Wang and Richard Ellison III for the work on Active Complex Event Processing
Congratulations to the winners,  and many thanks to all those who submitted award nominations, and to the EPTS award committee:  Alex Alves, Mani Chandy, Brenda Michelson, and Themis Palpanas.








DEBS 2011 tutorial day


The first day of the DEBS conference is traditionally the tutorials day.  There was a relatively large number of participants that attended this tutorials.  I have listened to one tutorial given by Nenad Stojanovic and Pedro Bizzaro on the event processing grand challenge set in Dagstuhl, May 2010.  I think that while all the ingredients are there, there still need to be some thinking about phrasing it in the grand challenge way, of an ambitious measurable goal.    The second tutorial I attended (as a captive audience) was the tutorial that I presented together with Tali Yatzkar-Haham.    I'll post this tutorial slides on slideshare, and write more about it within the next few days. Our tutorial covered the various issues of non-functional properties of event processing; there was some interesting discussion on several topics, like correctness, usability in general and transactional properties of event processing.    Tomorrow the main part of the conference starts.  More - later

Monday, July 11, 2011

Welcome to DEBS'11


After a year-long preparations, DEBS'11 is sailing out this morning.  I have arrived yesterday to the USA, and now writing from the  Westchester Marriott hotel in Tarrytown, which ACM selected as the hotel of residence for the conferences's participants.   The conference itself will take place in the IBM Research headquarter in Yorktown Heights.   


I am general chair of this conference, which meant a lot of work;  the local organization chairs who took care of all logistics - Bugra Gedik and Gabriela Jacques Da Silva, have done excellent job, so hopefully everything will be smooth.    


Looking forward to meet a lot of all friends, and some new ones,  it seems that the event processing research community is growing, and that there will be quite a lot of industrial participants as well.


Besides the "back office" work,  I'll have also some "front office" roles in the conference - starting today, I'll deliver a tutorial on non-functional properties of event processing, together with Tali Yatkar-Haham from my team, for her it will be the first participation in DEBS,  another person of the team Yagil Engel is planned to present here as well.     I'll also deliver one paper presentation in this conference (following Ella Rabinovich's thesis),  and have some speaking roles in the opening session and in the banquet.


Will be extremely busy days for me,  will also try to post some during the conference, I am sure it will also be covered by other bloggers (e.g. Paul Vincent).


More -later.