Learning Implicit and Explicit Control in Model Transformations by Example

Baki, Islem; Sahraoui, Houari; Cobbaert, Quentin; Masson, Philippe; Faunes, Martin

doi:10.1007/978-3-319-11653-2_39

Cited by 14 publications

(10 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, to our best knowledge, none of the contributions above can derive rules that require testing string attribute values or exploring a global context except for Baki et al [2014], where the result returned by a user-defined query can be used in rule conditions. We also noticed that most MTBE contributions illustrate their respective approaches using simplified transformation problems or toy examples without showing the complexity and correctness of the rules that can be learned.…”

Section: Example-based Model Transformationsmentioning

confidence: 98%

“…The algorithm is similar to the one described in Baki et al [2014] except it uses transformation traces to reduce the size of the search space. Unlike for the proposed approach, the standard GP algorithm aims to learn all the transformation program in a single step.…”

Section: Validationmentioning

confidence: 99%

“…First, the source and target metamodels are large and the transformation requires 26 rules. Second, most of the model elements have many attributes, which allows us to assess how our approach responds to the proliferation of attributes, a characteristic that hampered our previous contributions [Baki et al 2014].…”

Section: Experiments Settingmentioning

confidence: 99%

“…First, we assess the benefits of our learning process by comparing the fitness scores obtained for the transformation problem used in RQ3 (CL2RE) with two other samples obtained respectively by a random exploration of the search space, and a single-phase standard GP search as the described in Baki et al [2014] and Faunes et al [2013]. Second, we run the second phase of our approach several times with and without its adaptive features to assess the benefits brought by the proposed mutation strategies.…”

Section: Validationmentioning

confidence: 99%

“…The most-recent contribution to MTBE is the one of Faunes et al [2012Faunes et al [ , 2013 in which genetic programming (GP) is used to learn n-m transformation rules starting from source and target examples without the transformation traces. The approach is enhanced by Baki et al [2014] to learn the rule execution control.…”

Section: Related Workmentioning

confidence: 99%

See 4 more Smart Citations

Multi-Step Learning and Adaptive Search for Learning Complex Model Transformations from Examples

Baki

Sahraoui

2016

ACM Trans. Softw. Eng. Methodol.

Self Cite

View full text Add to dashboard Cite

Model-driven engineering promotes models as main development artifacts. As several models may be manipulated during the software-development life cycle, model transformations ensure their consistency by automating model generation and update tasks. However, writing model transformations requires much knowledge and effort that detract from their benefits. To address this issue, Model Transformation by Example (MTBE) aims to learn transformation programs from source and target model pairs supplied as examples. In this article, we tackle the fundamental issues that prevent the existing MTBE approaches from efficiently solving the problem of learning model transformations. We show that, when considering complex transformations, the search space is too large to be explored by naive search techniques. We propose an MTBE process to learn complex model transformations by considering three common requirements: element context and state dependencies and complex value derivation. Our process relies on two strategies to reduce the size of the search space and to better explore it, namely, multi-step learning and adaptive search. We experimentally evaluate our approach on seven model transformation problems. The learned transformation programs are able to produce perfect target models in three transformation cases, whereas precision and recall values larger than 90% are recorded for the four remaining cases. ACM Reference Format: Islem Baki and Houari Sahraoui. 2016. Multi-step learning and adaptive search for learning complex model transformations from examples.

show abstract

Section: Example-based Model Transformationsmentioning

confidence: 98%

Section: Validationmentioning

confidence: 99%

Section: Experiments Settingmentioning

confidence: 99%

Section: Validationmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 3 more Smart Citations

Multi-Step Learning and Adaptive Search for Learning Complex Model Transformations from Examples

Baki

Sahraoui

2016

ACM Trans. Softw. Eng. Methodol.

Self Cite

View full text Add to dashboard Cite

show abstract

Search‐based model transformations

Fleck

Troya

Wimmer

2016

J Software Evolu Process

View full text Add to dashboard Cite

Model transformations are an important cornerstone of model‐driven engineering, a discipline which facilitates the abstraction of relevant information of a system as models. The success of the final system mainly depends on the optimization of these models through model transformations. Currently, the application of transformations is realized either by following the apply‐as‐long‐as‐possible strategy or by the provision of explicit rule orchestrations. This implies two main limitations. First, the optimization objectives are implicitly hidden in the transformation rules and their orchestration. Second, manually finding the best orchestration for a particular scenario is a major challenge due to the high number of possible combinations. To overcome these limitations, we present a novel framework that builds on the non‐intrusive integration of optimization and model transformation technologies. In particular, we formulate the transformation orchestration task as an optimization problem, which allows for the efficient exploration of the transformation space and explication of the transformation objectives. Our generic framework provides several search algorithms and guides the user in providing a proper search configuration. We present different instantiations of our framework to demonstrate its feasibility, applicability, and benefits using several case studies. Copyright © 2016 John Wiley & Sons, Ltd.

show abstract