Tuesday, December 4, 2012

Some ICSE 2013 papers...


Papers from ICSE 2013:
Process
Not Going to Take This Anymore: Multi-objective Overtime Planning for Software Engineering Projects
Filomena Ferrucci, Mark Harman, Jian Ren, and Federica Sarro
(University of Salerno, Italy; University College London, UK)
NO RESPONSE

Search-Based SE
How to Effectively Use Topic Models for Software Engineering Tasks? An Approach Based on Genetic Algorithms
Annibale Panichella, Bogdan Dit, Rocco Oliveto, Massimiliano Di Penta, Denys Poshyvanyk, and Andrea De Lucia
(University of Salerno, Italy; College of William and Mary, USA; University of Molise, Italy; University of Sannio, Italy)
INFORMATION RETRIEVAL.
GA, single objective.

Search-Based Genetic Optimization for Deployment and Reconfiguration of Software in the Cloud
Sören Frey, Florian Fittkau, and Wilhelm Hasselbring
(Kiel University, Germany)
Problem: cloud deployment options (CDOs)
Simulator CDOSim can evaluate CDOs, e.g., regarding response times and costs.

Three objectives:
1-     cost refer to the total amount of monetary units owed to a cloud provider because of utilizing provided services.
2-     Response times refer to the average response times of the methods that are included in a workload profile.
3-     SLA violations indicate the number of method calls with response times that exceed a given threshold.
  
Custom algorithms based on NSGA-II.
Improvement by 60% over other tools.


Product Lines
Beyond Boolean Product-Line Model Checking: Dealing with Feature Attributes and Multi-features
Maxime Cordy, Pierre-Yves Schobbens, Patrick Heymans, and Axel Legay
(University of Namur, Belgium; INRIA, France)
Still a Verification method.
Multi-features: features that can appear several times in the same product, e.g. processing units.

On the Value of User Preferences in Search-Based Software Engineering: A Case Study in Software Product Lines
Abdel Salam Sayyad, Tim Menzies, and Hany Ammar
(West Virginia University, USA)

Reviewer #1:
 I suggest that the title be more direct in its description of the paper's content (i.e., make clear that the paper's focus is on evaluating the use of MEOAs in the SPL domain).

Reviewer #2:
The paper has a very nice survey of previous work on SBSE and multi objective SBSE that establishes the context in which the contributions of the paper are made. This is effectively a kind of "mini focused literature review” that focusses on previous work on multi objective SBSE. It is a small contribution
in itself.
The claims are very much overstated and need to be toned down. 
You can't say "who" when referring to an algorithm; they are not alive, no matter how "AI" this SBSE stuff might be :-)

Reviewer #3:
This is a very good paper. It reads like a page-turner
Also quite original and appealing is the idea of reducing the correctness criteria to just another dimension to optimize.

No comments:

Post a Comment