At long last, the EATCS has a YouTube channel. (Thanks to the EATCS Secretary office in Patras for setting it up and for making us enter the 21st century :-)) In due course, the videos of the invited talks and of the EATCS-Award-2013 presentation by Martin Dyer from ICALP 2013 will be available from that channel. For the moment, you can enjoy Erik Demaine's wonderful Presburger Award 2013 video.
Make sure that you subscribe to the channel to get updates on our postings.
The slides for Jon Kleinberg's inspiring EATCS lecture are available here.
Several people asked me for the slides and Jon agreed to make them public. Thanks Jon!
Conference photos taken at ICALP 2013 can be found here. Enjoy.
Tuesday, July 23, 2013
Thursday, July 18, 2013
CAV Award 2013
The CAV Award for 2013 has been presented to Kim G. Larsen, Paul Pettersson and Wang Yi for their seminal work on Uppaal, which is the foremost tool suite for the verification of real-time systems and the synthesis of real-time controllers. Congratulations to the awardees for receiving this well deserved accolade and to the award committee for making such an excellent choice!
If you are interested in using the tool for research and/or teaching, simply download it and read this tutorial by Frits Vaandrager. You will be up and running in no time.
Uppaal has its roots in a tool originally developed in Uppsala and described in the conference paper Automatic verification of real-time communicating systems by constraint-solving co-authored by Wang Yi, Paul Pettersson and Mads Daniels (Proceedings of FORTE 1994). Since then, Uppaal has been jointly developed by Kim G. Larsen's research group at Aalborg University and by the group led by Wang Yi at Uppsala University. In this period, Uppaal has become an industrial-strength tool for computer-aided verification of computing systems that has been applied to many case studies by several research groups in academia and industry. The efficiency of its computational engine has been improved greatly by theoretical and practical advances relying on highly non-trivial insights. Moreover, the tool now supports the analysis of quantitative extensions of timed automata, automatic model-based testing of real-time systems and the synthesis of controllers in the context of timed games. It is still under continuous development.
Overall, the Uppaal tool is a real success story for the research community working on automated verification of computer systems. As all long-term research and tool development efforts, the work on Uppaal and its applications is due to many gifted researchers and their students. The CAV Award 2013 cannot recognize them all, but the list of authors of the many papers stemming from the work on UPPAAL gives a hint of the magnitude of the research effort involved in building such a tool over a period of nearly 20 years.
If you are interested in using the tool for research and/or teaching, simply download it and read this tutorial by Frits Vaandrager. You will be up and running in no time.
Uppaal has its roots in a tool originally developed in Uppsala and described in the conference paper Automatic verification of real-time communicating systems by constraint-solving co-authored by Wang Yi, Paul Pettersson and Mads Daniels (Proceedings of FORTE 1994). Since then, Uppaal has been jointly developed by Kim G. Larsen's research group at Aalborg University and by the group led by Wang Yi at Uppsala University. In this period, Uppaal has become an industrial-strength tool for computer-aided verification of computing systems that has been applied to many case studies by several research groups in academia and industry. The efficiency of its computational engine has been improved greatly by theoretical and practical advances relying on highly non-trivial insights. Moreover, the tool now supports the analysis of quantitative extensions of timed automata, automatic model-based testing of real-time systems and the synthesis of controllers in the context of timed games. It is still under continuous development.
Overall, the Uppaal tool is a real success story for the research community working on automated verification of computer systems. As all long-term research and tool development efforts, the work on Uppaal and its applications is due to many gifted researchers and their students. The CAV Award 2013 cannot recognize them all, but the list of authors of the many papers stemming from the work on UPPAAL gives a hint of the magnitude of the research effort involved in building such a tool over a period of nearly 20 years.
Monday, July 15, 2013
ICALP 2013 (Part 3): Award sessions
The best paper award session at ICALP 2013 was held last Monday before the welcome reception (which included excellent, and plentiful, finger food and wine). Mark Bun, John Fearnley and Dominik Pajak
delivered very good presentations of the award-winning papers. I hope
that you will check out their papers. In the case of the paper by Mark
and Justin Thaler, you can also read this expository blog post.
The EATCS Award and the Presburger Award were delivered last Tuesday in a joint ceremony that also included the award of honorary doctorates to Josef Gruska and Juris Hartmanis. (See this earlier post of mine for some more information on the ceremony for the honorary doctorates.)
The master of ceremony for the award session was Paul Spirakis. First Martin Dyer received the EATCS Award from Friedhelm Meyer auf der Heide and delivered a presentation entitled Counting ain't easy. Martin started by citing the following fragment of a rhyme from the Winnie the Pooh books:
Cheers for Pooh!
(For who?)
For Pooh -
(Why, what did he do?)
I thought you knew
He then proceeded to give an accessible historical overview of what he did do, covering essentially all the work mentioned in the laudatio for the award.
Last, but by no means least, Tony Kucera presented the Presburger Award 2013 to Erik Demaine. Unfortunately, Erik could not be with us in Riga, but we had a virtual presentation of the award to him via Skype. Moreover, Erik produced an excellent video of a presentation that we could play and enjoy at the conference. Rather than attempting to summarize Erik's inspirational talk, I will simply limit myself to letting him speak for himself.
The EATCS Award and the Presburger Award were delivered last Tuesday in a joint ceremony that also included the award of honorary doctorates to Josef Gruska and Juris Hartmanis. (See this earlier post of mine for some more information on the ceremony for the honorary doctorates.)
The master of ceremony for the award session was Paul Spirakis. First Martin Dyer received the EATCS Award from Friedhelm Meyer auf der Heide and delivered a presentation entitled Counting ain't easy. Martin started by citing the following fragment of a rhyme from the Winnie the Pooh books:
Cheers for Pooh!
(For who?)
For Pooh -
(Why, what did he do?)
I thought you knew
He then proceeded to give an accessible historical overview of what he did do, covering essentially all the work mentioned in the laudatio for the award.
Last, but by no means least, Tony Kucera presented the Presburger Award 2013 to Erik Demaine. Unfortunately, Erik could not be with us in Riga, but we had a virtual presentation of the award to him via Skype. Moreover, Erik produced an excellent video of a presentation that we could play and enjoy at the conference. Rather than attempting to summarize Erik's inspirational talk, I will simply limit myself to letting him speak for himself.
Sunday, July 14, 2013
ICALP 2013 (Part 2): Invited talks
ICALP 2013 in Riga featured five invited presentations and a special EATCS lecture to celebrate the 40th edition of the ICALP conference.
The scientific program for the conference was preceded by short presentations by the rector of the University of Latvia and by Rusins Freivalds. The rector welcomed the ICALP participants, gave us some interesting information about the University of Latvia and wished us long coffee breaks. Rusins discussed the unity of science, and reminded us that, even at the time of the iron curtain, there was one Computer Science.
The conference proper was kicked off on Monday, 8 July, by an invited talk delivered by the Liverpool-bound Paul Spirakis. Paul's talk was entitled A Guided Tour in Random Intersection Graphs. Random Intersection Graphs are random graphs in which there is a universe M of labels and each one of the vertices selects a random subset of M. Two vertices are connected if, and only if, their corresponding subsets of labels intersect. Random intersection graphs were introduced by M. Karonski, E.R. Sheinerman and K.B. Singer-Cohen and have several applications, as well as a rich theory. Paul's talk provided a survey of the main results on the topic obtained by his research group on combinatorial problems over random intersection graphs, such as independent sets, Hamiltonian cycles, colouring, maximum cliques, expansion and random walks. Paul closed the talk by saying that this model is an excellent area of study for PhD students in TCS.
Tuesday's invited address was delivered by Daniel Marx. Daniel's talk has three chapters (his words) and was entitled The Square Root Phenomenon in Planar Graphs. Its starting point was the observation that most of the classic NP-hard problems remain NP-hard when restricted to planar graphs, and only exponential-time algorithms are known for the exact solution of these planar problems. However, in many cases, the exponential-time algorithms on planar graphs are significantly faster than the algorithms for general graphs: for various problems on planar graphs, one often sees a square root appearing in the exponent of the running time of the best algorithms for their solution. Daniel told his audience that, by now, we have a good understanding of why this square root appears: most of these algorithms rely on the notion of treewidth and its relation to grid minors in planar graphs. Daniel also argued that, under the Exponential Time Hypothesis, one can show that these algorithms are essentially best possible, and therefore the square root must appear in the running time.
In passing, Daniel contributed also one paper to ICALP Track A and one to ICALP Track B!
Susanne Albers delivered the invited talk on Wednesday, 10 July, on Recent Advances for a Classical Scheduling Problem. In her talk, Susanne revisited the classic on-line makespan minimization problem, which has been studied since the 1960s. After presenting the classic results on this problem, starting from Graham's 1966 List algorithm and its competitive analysis, she surveyed recent research on settings in which an online algorithm is given extra information or power while processing a job sequence.
The scientific programme on Thursday, 11 July, started with an invited talk by Orna Kupferman, who gave the only invited address that could be readily classified as belonging to ICALP Track B. Orna's presentation dealt with Formalizing and Reasoning about Quality. Traditionally, formal approaches to the verification of reactive systems are boolean in nature: either a system satisfies its specification or it doesn't. In case a system does not meet its specification, one expects a good verification framework to provide a counter-example, that is, a reason why the system is not correct. Orna and her co-authors have generalized formal specification and verification methods to address the quality of systems. In her talk, Orna introduced the linear temporal logic LTL[F], where F is a set arbitrary functions over the interval [0, 1]. Formulae in LTL[F] are interpreted over computations consisting of sequences of atomic propositions. The satisfaction value of an LTL[F] formula is a number between 0 and 1 that describes how well a computation satisfies a formula. The logic generalizes traditional LTL with the functions in F; examples of functions that might be in F are the maximum or minimum between the satisfaction values of subformulas (these are the quantitative counterparts of boolean OR and AND, respectively), their product and their average. In her talk, Orna showed us how to generalize classic decision problems in formal methods, such as satisfiability, model checking and synthesis, to search and optimization problems in the quantitative setting. This is achieved by means of an extension of the automata-theoretic approach to LTL to the setting of LTL[F].
Before the conference dinner, Jon Kleinberg delivered a special EATCS lecture to celebrate the 40th ICALP. Jon gave an inspiring and very articulate talk entitled Algorithms, Networks, and Social Phenomena. The talk discussed the development of computational models for systems involving social networks and large human audiences. In particular, Jon focused on how information spreads through such systems, and the ways in which this spread is affected by the underlying network structure. Jon said a few times that, despite having so much data at our disposal, we still do not understand human behaviour. However, IMHO, the work by Jon and his coworkers is shedding some light on some aspects of our behaviour.
Peter Widmayer delivered the last invited talk on Friday, 12 July. His presentation was entitled To Be Uncertain Is Uncomfortable, But to Be Certain Is Ridiculous, and was accessible and well paced. The starting point of Peter's talk was a "Socratic dialogue" between a statistical physicist and himself. Traditionally, in combinatorial optimization one assumes that an input instance is given with absolute certainty. The goal is then to find an optimum solution for the given instance. In contrast, as the statistical physicist would argue, input data are in reality uncertain, noisy and inaccurate. As a consequence, an optimum solution to a combinatorial optimization problem might not be meaningful in practice. (For example, the shortest path to our work place we computed yesterday evening might not be usable this morning because of changed traffic conditions.) Peter advocated the development of algorithms that find "meaningful" solutions in the presence of uncertain inputs, proposed an approach towards reaching this goal. and argued that it leads to good solutions in the real world.
Videos of these talks (with the exception of Kleinberg's talk, which, as far as I know, was not recorded) will be available in due course.
The scientific program for the conference was preceded by short presentations by the rector of the University of Latvia and by Rusins Freivalds. The rector welcomed the ICALP participants, gave us some interesting information about the University of Latvia and wished us long coffee breaks. Rusins discussed the unity of science, and reminded us that, even at the time of the iron curtain, there was one Computer Science.
The conference proper was kicked off on Monday, 8 July, by an invited talk delivered by the Liverpool-bound Paul Spirakis. Paul's talk was entitled A Guided Tour in Random Intersection Graphs. Random Intersection Graphs are random graphs in which there is a universe M of labels and each one of the vertices selects a random subset of M. Two vertices are connected if, and only if, their corresponding subsets of labels intersect. Random intersection graphs were introduced by M. Karonski, E.R. Sheinerman and K.B. Singer-Cohen and have several applications, as well as a rich theory. Paul's talk provided a survey of the main results on the topic obtained by his research group on combinatorial problems over random intersection graphs, such as independent sets, Hamiltonian cycles, colouring, maximum cliques, expansion and random walks. Paul closed the talk by saying that this model is an excellent area of study for PhD students in TCS.
Tuesday's invited address was delivered by Daniel Marx. Daniel's talk has three chapters (his words) and was entitled The Square Root Phenomenon in Planar Graphs. Its starting point was the observation that most of the classic NP-hard problems remain NP-hard when restricted to planar graphs, and only exponential-time algorithms are known for the exact solution of these planar problems. However, in many cases, the exponential-time algorithms on planar graphs are significantly faster than the algorithms for general graphs: for various problems on planar graphs, one often sees a square root appearing in the exponent of the running time of the best algorithms for their solution. Daniel told his audience that, by now, we have a good understanding of why this square root appears: most of these algorithms rely on the notion of treewidth and its relation to grid minors in planar graphs. Daniel also argued that, under the Exponential Time Hypothesis, one can show that these algorithms are essentially best possible, and therefore the square root must appear in the running time.
In passing, Daniel contributed also one paper to ICALP Track A and one to ICALP Track B!
Susanne Albers delivered the invited talk on Wednesday, 10 July, on Recent Advances for a Classical Scheduling Problem. In her talk, Susanne revisited the classic on-line makespan minimization problem, which has been studied since the 1960s. After presenting the classic results on this problem, starting from Graham's 1966 List algorithm and its competitive analysis, she surveyed recent research on settings in which an online algorithm is given extra information or power while processing a job sequence.
The scientific programme on Thursday, 11 July, started with an invited talk by Orna Kupferman, who gave the only invited address that could be readily classified as belonging to ICALP Track B. Orna's presentation dealt with Formalizing and Reasoning about Quality. Traditionally, formal approaches to the verification of reactive systems are boolean in nature: either a system satisfies its specification or it doesn't. In case a system does not meet its specification, one expects a good verification framework to provide a counter-example, that is, a reason why the system is not correct. Orna and her co-authors have generalized formal specification and verification methods to address the quality of systems. In her talk, Orna introduced the linear temporal logic LTL[F], where F is a set arbitrary functions over the interval [0, 1]. Formulae in LTL[F] are interpreted over computations consisting of sequences of atomic propositions. The satisfaction value of an LTL[F] formula is a number between 0 and 1 that describes how well a computation satisfies a formula. The logic generalizes traditional LTL with the functions in F; examples of functions that might be in F are the maximum or minimum between the satisfaction values of subformulas (these are the quantitative counterparts of boolean OR and AND, respectively), their product and their average. In her talk, Orna showed us how to generalize classic decision problems in formal methods, such as satisfiability, model checking and synthesis, to search and optimization problems in the quantitative setting. This is achieved by means of an extension of the automata-theoretic approach to LTL to the setting of LTL[F].
Before the conference dinner, Jon Kleinberg delivered a special EATCS lecture to celebrate the 40th ICALP. Jon gave an inspiring and very articulate talk entitled Algorithms, Networks, and Social Phenomena. The talk discussed the development of computational models for systems involving social networks and large human audiences. In particular, Jon focused on how information spreads through such systems, and the ways in which this spread is affected by the underlying network structure. Jon said a few times that, despite having so much data at our disposal, we still do not understand human behaviour. However, IMHO, the work by Jon and his coworkers is shedding some light on some aspects of our behaviour.
Peter Widmayer delivered the last invited talk on Friday, 12 July. His presentation was entitled To Be Uncertain Is Uncomfortable, But to Be Certain Is Ridiculous, and was accessible and well paced. The starting point of Peter's talk was a "Socratic dialogue" between a statistical physicist and himself. Traditionally, in combinatorial optimization one assumes that an input instance is given with absolute certainty. The goal is then to find an optimum solution for the given instance. In contrast, as the statistical physicist would argue, input data are in reality uncertain, noisy and inaccurate. As a consequence, an optimum solution to a combinatorial optimization problem might not be meaningful in practice. (For example, the shortest path to our work place we computed yesterday evening might not be usable this morning because of changed traffic conditions.) Peter advocated the development of algorithms that find "meaningful" solutions in the presence of uncertain inputs, proposed an approach towards reaching this goal. and argued that it leads to good solutions in the real world.
Videos of these talks (with the exception of Kleinberg's talk, which, as far as I know, was not recorded) will be available in due course.
Friday, July 12, 2013
ICALP 2013 (Part 1)
ICALP 2013 has just finished. It has been an action packed conference and there has been next to no breathing space. I think that the conference was a great success and the organizers deserve our most heartfelt thanks for all their work.
In case you do not know, the conference has being held in Riga at the University of Latvia. This has been the first time that ICALP has been organized in a country of the former Soviet Union and this was the 40th edition of the ICALP conference. First of all, kudos to the local organizers, who did their very best to make the conference a festive occasion and a very pleasant experience for all the attendees. The city of Riga is very pretty and all ICALP participants had a very warm welcome. The University of Latvia was also enrolling a new batch of students during ICALP and this meant that there was a continuous flow of young and happy-looking people on the university premises. This made the main hall of the university a very lively place to be. We were also blessed with sunny and warm weather, which also helped (especially coming from rainy Iceland).
According to the data presented by Agnis Škuškovniks, on behalf of the organizing committee, to the EATCS council and to the EATCS general assembly, there were 193 registered participants in ICALP 2013, of which 67 are students. Including the local participants, the six invited speakers and the two recipients of the honorary doctorates awarded on Tuesday (Josef Gruska and Juris Hartmanis), 217 people attended ICALP 2013. The pre-conference workshops were attended by 102 participants, which is a very healthy number.
The honorary doctorates were an excellent addition to the standard session devoted to the EATCS Awards. Gruska and Hartmanis delivered lucid and inspirational presentations. It was truly awesome to see Juris Hartmanis deliver an off-the-cuff speech on how he left Latvia and ended up at Cornell as chair of the newly-funded CS department. At 85, he is very articulate and sharp. I had the pleasure of discussing several subjects with him during a very pleasant dinner arranged by the organizers. He is still a truly inspirational figure.
Here are some quick EATCS- and ICALP-related news.
In a subsequent post, which I will write when I get home, I will try to discuss some of the many scientific highlights at ICALP 2013.
To conclude, at the conference I heard that Paul Spirakis is moving to the University of Liverpool. Good luck to Paul!
In case you do not know, the conference has being held in Riga at the University of Latvia. This has been the first time that ICALP has been organized in a country of the former Soviet Union and this was the 40th edition of the ICALP conference. First of all, kudos to the local organizers, who did their very best to make the conference a festive occasion and a very pleasant experience for all the attendees. The city of Riga is very pretty and all ICALP participants had a very warm welcome. The University of Latvia was also enrolling a new batch of students during ICALP and this meant that there was a continuous flow of young and happy-looking people on the university premises. This made the main hall of the university a very lively place to be. We were also blessed with sunny and warm weather, which also helped (especially coming from rainy Iceland).
According to the data presented by Agnis Škuškovniks, on behalf of the organizing committee, to the EATCS council and to the EATCS general assembly, there were 193 registered participants in ICALP 2013, of which 67 are students. Including the local participants, the six invited speakers and the two recipients of the honorary doctorates awarded on Tuesday (Josef Gruska and Juris Hartmanis), 217 people attended ICALP 2013. The pre-conference workshops were attended by 102 participants, which is a very healthy number.
The honorary doctorates were an excellent addition to the standard session devoted to the EATCS Awards. Gruska and Hartmanis delivered lucid and inspirational presentations. It was truly awesome to see Juris Hartmanis deliver an off-the-cuff speech on how he left Latvia and ended up at Cornell as chair of the newly-funded CS department. At 85, he is very articulate and sharp. I had the pleasure of discussing several subjects with him during a very pleasant dinner arranged by the organizers. He is still a truly inspirational figure.
Here are some quick EATCS- and ICALP-related news.
- ICALP 2014 will be held in Copenhagen, Denmark, immediately after SWAT 2013. It will be a four-day ICALP and the general chair for the conference is Thore Husfeldt.
- ICALP 2015 will be held in Kyoto, Japan, and will be co-located with LICS 2015. Kazuo Iwama is the ICALP 2015 general chair. This will be the first ever ICALP outside Europe.
- The EATCS will soon announce and EATCS Fellows programme. The first deadline for nominations will be at the end of this year and we plan to make the first announcement of fellows at ICALP 2014 in Copenhagen.
- The EATCS will begin a new EATCS Young Researcher School Series from 2014. The first school in the series will be organized by Tony Kucera in a beautiful location in Moravia in late July/early August. The theme of the school will be "Automata, Logic and Games".
- Orna Kupferman gave a thought-provoking talk on "The Gender Challenge in TCS" at the EATCS General Assembly. It really got the audience thinking about this important matter.
In a subsequent post, which I will write when I get home, I will try to discuss some of the many scientific highlights at ICALP 2013.
To conclude, at the conference I heard that Paul Spirakis is moving to the University of Liverpool. Good luck to Paul!
Monday, July 08, 2013
Associate and/or assistant professor position in the Language Based Technology section of DTU
I have been asked to advertise this position, which might be of interest to several of my readers and their graduating PhD students/postdocs.
The Language Based Technology section of DTU Applied Mathematics and Computer Science has a new opening for associate and/or assistant professors.
The Language Based Technology section of DTU Applied Mathematics and Computer Science has a new opening for associate and/or assistant professors.
The date for applictions is on October 1st and full details about the call and how to apply are available at
and more information about the Language Based Technology section is available at
We are hoping to attract a brilliant associate or assistant professor to complement our research activities in the
modelling, analysis and realisation of systems using language-based techniques and tools.
Our current expertices include static analysis and model checking
and take place in a number of research centres and research projects
(MT-LAB, IDEA4CPS, TREsPASS, FutureID, SESAMO, PaPP).
We are interested in concurrent, distributed and mobile systems
where reliability and predictability are essential and hence properties
related to safety, security and performance are of key interest; the
systems may include considerations of socio-technical
nature within Systems of Systems.
Subscribe to:
Posts (Atom)