research and more

14
Research and more

Upload: evelina-jara

Post on 31-Dec-2015

18 views

Category:

Documents


1 download

DESCRIPTION

Research and more. Who am I?. B.Sc. Math/Computer Science Hebrew University of Jerusalem, Israel M.Sc. Computer Science Hebrew University of Jerusalem, Israel Ph.D. in Science (Computer Science) Katholieke Universiteit Leuven, Belgium Postdoc at Ecole Polytechnique, France LaQuSo. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Research and more

Research and more

Page 2: Research and more

/ name of department PAGE 204/19/23

Who am I?

• B.Sc. Math/Computer Science • Hebrew University of Jerusalem, Israel

• M.Sc. Computer Science• Hebrew University of Jerusalem, Israel

• Ph.D. in Science (Computer Science)• Katholieke Universiteit Leuven, Belgium

• Postdoc at Ecole Polytechnique, France

• LaQuSoBackground – Recent research – More…

Page 3: Research and more

/ name of department PAGE 304/19/23

Background (1997-2004)

• Theses: Termination analysis of logic programs

• Abstraction techniques in program analysis• But also:

• Refactoring in LP [Demoen/Schrijvers/S]

• View-usability for queries with aggregation [Cohen/Nutt/Sagiv/S]

• XML query languages [Cohen/Kanza/Kogan/Nutt/Sagiv/S]

Background – Recent research – More…

Page 4: Research and more

/ name of department PAGE 404/19/23

Termination analysis: What is this about?

• Undecidable for “interesting languages”• Necessary and sufficient conditions

[Dershowitz/Lindenstrauss/Sagiv/S, De Schreye/S]

• Sufficient conditions that can be verified automatically [Idem]

− Usually: approximations of necessary and sufficient conditions

− Often: abstractions or transformations • Interesting decidable sublanguages [Mesnard/S]

Background – Recent research – More…

Page 5: Research and more

/ name of department PAGE 504/19/23

Background (2005-2006)

• Petri nets (modelling and analysis):• Resources • Adaptivity

• Termination analysis:• Decidable sublanguages [Mesnard/S]

• New approach using TRS [Schneider-Kamp/Giesl/S/Thiemann]

• But also: • BKE with μCRL [Bloom/Grote/Mauw/S]

• Code stability [Huizing/Kuiper/Punter/S]

• Popular papers [Heck/S]

Page 6: Research and more

/ name of department PAGE 604/19/23

Resources in Petri nets

• Durable and identical resources• Soundness

• Whatever the system does…• There is a way to complete the process:

and release all resources

• Additional resources should not disturb soundness.

• Is there a minimal number of resources that ensure soundness for any number of cases?

• [Hee/S/Sidorova/Voorhoeve]

Background – Recent research – More…

Page 7: Research and more

/ name of department PAGE 704/19/23

Adaptivity and workflow

Classical workflow:

structure is fixed during the design

phase

Adaptive workflow: structure can be modified in a structured way

Ad-hoc workflow: structure can be modified arbitrarily

[vd Aalst et al. 1999]

Structured way: parts of the process can be defined or extended later on…

1. No formal semantics

Verification is challenging

2. No reuse of existing WFMS!Background – Recent research – More…

Page 8: Research and more

/ name of department PAGE 804/19/23

Our solution: Nested Nets

Token is a net!

…Level 2

Level 0

Level 1

Analysis is possible using abstractions [Hee/Oanea/S/Sidorova/Voorhoeve/Lomazova]

Application: medical

protocols[Hee/Schonenberg/S/Sidorova/vdWerf]

Background – Recent research – More…

Page 9: Research and more

/ name of department PAGE 904/19/23

Recent research (2007-2008)

• Petri nets (modelling and analysis):• History (global vs. local) • History + adaptivity [Hee/Schonenberg/S/Sidorova/vdWerf]

• Mining [vdWerf/vDongen/Hee/Hurkens/S]

• Compositionality [Massuthe/S/Sidorova/Wolf]

• Improved termination analysis [Schneider-Kamp/Giesl/S/Thiemann]

• Framework for automatic verification [Raeds/Petković/S/Somers]

• Metrics and Reverse engineering [Petković/vdBrand/S/Korshunova]

Background – Recent research – More…

Page 10: Research and more

/ name of department PAGE 1004/19/23

Recent research: History

• Modelling log-based decisions.

• Twee models:• Global history (log is accessible for everybody)• Token history (each token has its own past)

• Translating history models to classical Petri nets

• Analysis• [Hee/S/Sidorova/vd Aalst, Hee/S/Sidorova]

Background – Recent research – More…

Page 11: Research and more

/ name of department PAGE 1104/19/23

Recent research: Termination

• LP TRS; analysis of a TRS

• Existing translations:• Restricted to a subclass of LP• Often: LP terminates, TRS – does not!

• New translation• New form of rewriting• Provably more powerful than previous transformation-

based approaches

• [Schneider-Kamp/Giesl/S/Thiemann]

Background – Recent research – More…

Page 12: Research and more

/ name of department PAGE 1204/19/23

Research interests

• Static analysis using abstraction techniques

• Maintainability

• Termination analysis

Background – Recent research – More…

Page 13: Research and more

/ name of department PAGE 1304/19/23

Even more

• LaQuSo cases: • ABP 2, Almende, Braunschweig, Cepo, Cordys, De Lage

Landen, De Lage Landen 2, Dynafix, M2M, NBG, NBG 2, Oce.

• Code analysis• Architecture reviews• Literature study• Requirements elicitation • Certification

Background – Recent research – More…

Page 14: Research and more

/ name of department PAGE 1404/19/23

Example of a LaQuSo case