Special Issue of the Automated Software Engineering Journal
Next Generation Search-Based Software Engineering: Insights from Search and Data Mining

Guest Editors: Marouane Kessentini, University of Michigan, Dearborn, MI, USA marouane@umich.edu
Tim Menzies, North Carolina State University, NC, USA tjmenzie@ncsu.edu

CALL FOR PAPERS

The concept of search has been explored in many fields such as Search-Based Software Engineering and Data Mining (e.g. one way to characterize data mining is a heuristic search though a large space of possible generalizations). That is, there is a strong theoretical connection between the goals of Search-Based SE and data mining. Hence, many papers have explored combinations of data mining and SBSE, using a range of techniques. For example:

• When tuning control parameters for a genetic algorithm, data miners can learn good tunings;

• When mutating current solutions, data miners can be applied to learn especially effective mutations;

• When explaining the results of search-based SE, data miners can summarize the Pareto frontier;

• When niching particle swam optimization, clustering algorithms can be inter-levered with the search;

• When the fitness function is difficult to define and/or expensive, data mining can be used to learn how the solutions can be evaluated.

The goal of this special issue is to understand the cost/benefit trade offs in combining Search-based SE and data mining. There are many questions in this field including (but not restricted to) the following:

• Data mining takes time and CPU so when is it worthwhile to add to the CPU cost of the search?

• When is it useful to adjust the control parameters of a search-based SE tool, using data mining?

• As a search algorithm walks through a problem space, how do we recognize when prior data mining results need to be revisited?

• If we must use data mining to summarize the Pareto frontier, should we replace (to some extent) the search algorithms with data mining algorithms?

• Can data mining be used to evaluate generated solutions or reduce the cost of evaluating them?

SCOPE

This special issue encourages high quality submissions on all aspects of combing data mining and search-based software engineering. Note that the evaluation section of all such submissions must use at least some software engineering case studies such as (but are not limited to) :

• Project Management and Organization

• Effort and Defect Prediction

• Requirements Engineering

• Developing Dynamic Service-Oriented Systems

• Configuring Cloud-Based Architectures

• Enabling Self-healing/Self-optimizing Systems

• Creating Recommendation Systems to Support Software Development

• Software Security

• System and Software Integration

• Test Data Generation

• Regression Testing Optimization

• Network Design and Monitoring

• Software Maintenance, Program Repair, Refactoring and Transformation

SUBMISSION GUIDELINES

Submission of a manuscript implies that the work described has not been published before. A submission extended from a previous conference version has to contain at least 30% new material. Authors are requested to attach to the submitted paper their previously published articles and an explanation of the novel contributions made in the journal version. Papers should be submitted to the special issue through the Editorial Manager https://www.editorialmanager.com/ause/, selecting the article type "Special Issue: Search-based Software Engineering and Predictive Modelling". Formatting templates can be found in the following link: http://www.springer.com/computer/ai/journal/10515?detailsPage=pltci_1060168

IMPORTANT DATES

• 1 April 2015 full paper submission deadline

• 15 June 2015 authors notification

• 1 September 2015 author revisions

• 15 November 2015 final notification

• 15 December 2015 final manuscript