SATToSE 2011 Presentations

<<Back to main page

Talks are listed by sessions

Tools session

The 101companies software corpus (slides1) (slides2) (slides3)
by Ralf Lämmel, Thomas Schmorleiz, and Andrei Varanovich (University of Koblenz-Landau)

5604369122_8ca812ff5b_m.jpg 5602560807_c71ea59689_m.jpg 5602566889_5ba786a5a2_m.jpg

Generation of corrective actions to code design problems
by Sergio Castro (Université catolique de Louvain)
5602557887_810123a647_m.jpg
In this presentation we will show Heal, a framework for helping architects and programmers to semi-automatically diagnose and correct design problems in source code.

Understanding evolution

Co-Evolution of Eclipse Plug-in Architecture and its Third-party Plug-ins (slides)
by John Businge (Eindhoven University of Technology)
5604350104_fb63f99d8c_m.jpg
We study the evolution of third-party plug-ins with respect to their dependencies on the Eclipse platform over time. We also study the evolution of Eclipse platform APIs and identify some of the characteristics or properties baring an effect on the evolution of the third-party plug.ins over time.

Software Evolution Towards Model-Centric Runtime Adaptivity (slides)
by Mahdi Derakhshanmanesh (University of Koblenz-Landau)
5602569929_387227f2af_m.jpg
Presentation of results of a joint work by Mehdi Amoui, Mahdi Derakhshanmanesh, Jürgen Ebert, Ladan Tahvildari (University of Waterloo, University of Koblenz-Landau).

Empirical analysis of the evolution of open source software ecosystems (slides)
by Tom Mens (University of Mons)
5603147620_fb29e8b823_m.jpg

Process Mining Software Repositories
by Alexander Serebrenik (Eindhoven University of Technology)

Challenges with the aggregation of software metrics (slides)
by Bogdan Vasilescu (Eindhoven University of Technology)
5603150638_47d5ecb258_m.jpg
Software metrics are usually defined at micro-level (method, class, package), while studying the maintainability and evolution of a software system typically requires insights at macro- (system-) level. In this presentation we explore some of the challenges that traditional (sum, mean etc.) or more recent (econometric inequality indices) aggregation techniques bring out.

SLE

A conceptual and generative approach to data-intensive program evolution (slides)
by Anthony Cleve (University of Namur)
What you (re)generate does not need to be written/transformed. In this talk, we present an approach according to which data access APIs are generated from (evolving) schema mappings, so that schema evolutions can be (partly) propagated as API regeneration.

Logic-based program transformation in symbiosis with Eclipse (slides)
by Coen De Roover (Vrije Universiteit Brussel)
5602562045_6558d9cc63_m.jpg

VM Endoscopy — Would you trust a blind surgeon? (slides)
by Toon Verwaest (Software Composition Group)
5602571117_545949ca7f_m.jpg
Current programming languages are so rigid that their users reach out to external DSLs to sweep the required boilerplate code under the carpet. I present a model of flexible object layouts, first-class slots and class modifications to enable the evolution of object-oriented languages and their runtime.

Dynamic software update with ActiveContext (slides)
by Erwann Wernli (University of Bern)
5602572221_b139236565_m.jpg
ActiveContext is a mechanism for dynamic software evolution aimed at dynamic languages which support proper state transfer and prevents old code on execution state to access update data structure in an usafe way.

Consistency management

A petri net execution model for managing inconsistencies in dynamically changing programs (slides)
by Nicolas Cardozo (Université catolique de Louvain)
5603143506_19ebe9f5cb_m.jpg
Checking (model) constraints for software evolutions is usually a cumbersome task. This task becomes even more critic when the software is to dynamically change at run-time. This presentation shows exploration use of petri nets to ease the verification of such constrains.

FB : A Fast-Backward Planner to resolve design-model inconsistencies (slides)
by Jorge Pinna Puissant (UMONS)
5602564801_634e04bfdc_m.jpg
Automated Planning is a technique coming from the Artificial Intelligence domain. This technique allows the generation of possible resolution plans without the need of manually writing resolution rules. FB (Fast-Backward Planner) is an automated planning approach developed to resolve model inconsistencies. FB is still under development and we will present the results of this approach.

Assessing SAT solvers for design model inconsistency resolution - a first study (slides)
by Ragnhild Van Der Straeten (Vrije Universiteit Brussel)
5602564035_0b49f9707b_m.jpg

Empirical techniques

Source Code Archeology using Logic Program Queries across Version Repositories (slides)
by Andy Kellens (Vrije Universiteit Brussel)
5602567949_440173db47_m.jpg

Dependency Analysis of Software Ecosystems
by Mircea Lungu (University of Bern)
5602573231_1d0412cf05_m.jpg
Dependency analysis usually happens at the individual system level. However, systems do not exist by themselves, but rather, they co-exist in larger software ecosystems. I will present techniques for dependency analysis for entire software ecosystems.

API usage analysis (slides)
by Ekaterina Pek (University of Koblenz-Landau)

<<Back to main page