Model Repair with Quality-Based Reinforcement Learning.

Iovino, Ludovico; Rodríguez, A.; Rutle, Adrian; Heldal, Rogardt

doi:10.5381/jot.2020.19.2.a17

Cited by 14 publications

(11 citation statements)

References 8 publications

(11 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Human Interaction Several researchers (Reder & Egyed 2012;Nassar et al 2017;Ludovico et al 2020) have argued that the user should play a leading role in the model repair process. We distinguish several ways of human interaction, coarsely grouped into upfront (performed before the repair invocation) and interactive (performed during the repair) measures.…”

Section: Search-based Solutionsmentioning

confidence: 99%

“…Learning, Incrementality and Invocation "Learning is the ability of a program to improve its performance on a given task over time" (Russell & Norvig 2010). It appears to be promising enhancement for improving the performance of search-based approaches (Barriga et al 2018) and can help to identify hidden policies and user preferences (Barriga et al 2020;Ludovico et al 2020).…”

Section: Search-based Solutionsmentioning

confidence: 99%

See 1 more Smart Citation

Multi-Model Evolution through Model Repair.

Stünkel¹,

König²,

Rutle³

et al. 2021

JOT

Self Cite

View full text Add to dashboard Cite

Design and development of complex software systems usually comprises multiple inter-related models, i.e. abstract representations of certain aspects of the underlying system. The relations between these models induce global consistency conditions which the models collectively must fulfill. At the same time, these models are subject to frequent changes, and as a result, maintaining their global consistency over time becomes an important issue in model management in general and Model-Driven Software Engineering in particular.In this paper, we present a comprehensive feature model providing an overview of the current state of the art of model management. In this feature model, we further identify the central role of model repair as an implementation pattern for (multi-)model evolution.

show abstract

Section: Search-based Solutionsmentioning

confidence: 99%

Section: Search-based Solutionsmentioning

confidence: 99%

Multi-Model Evolution through Model Repair.

Stünkel¹,

König²,

Rutle³

et al. 2021

JOT

Self Cite

View full text Add to dashboard Cite

show abstract

“…It is important to note that a labelled dataset is useful for testing clustering techniques since there are quality metrics that require the ground truth to be computed (e.g., Rand Index, NMI, AMI, etc). On the other hand, reinforcement learning approaches have been used in the modelling domain to apply automatic repairs [26].…”

Section: Machine Learning and Datasetsmentioning

confidence: 99%

ModelSet: a dataset for machine learning in model-driven engineering

López

Izquierdo²,

Cuadrado

2021

Softw Syst Model

View full text Add to dashboard Cite

The application of machine learning (ML) algorithms to address problems related to model-driven engineering (MDE) is currently hindered by the lack of curated datasets of software models. There are several reasons for this, including the lack of large collections of good quality models, the difficulty to label models due to the required domain expertise, and the relative immaturity of the application of ML to MDE. In this work, we present ModelSet, a labelled dataset of software models intended to enable the application of ML to address software modelling problems. To create it we have devised a method designed to facilitate the exploration and labelling of model datasets by interactively grouping similar models using off-the-shelf technologies like a search engine. We have built an Eclipse plug-in to support the labelling process, which we have used to label 5,466 Ecore meta-models and 5,120 UML models with its category as the main label plus additional secondary labels of interest. We have evaluated the ability of our labelling method to create meaningful groups of models in order to speed up the process, improving the effectiveness of classical clustering methods. We showcase the usefulness of the dataset by applying it in a real scenario: enhancing the MAR search engine. We use ModelSet to train models able to infer useful metadata to navigate search results. The dataset and the tooling are available at https://figshare.com/s/5a6c02fa8ed20782935c and a live version at http://modelset.github.io.

show abstract

“…Since PARMOREL is extensible, users can plug different tools to obtain these rewards. In this paper, we work with a quality metrics tool [Iovino et al 2020] inspired by the quality model proposed in [Basciani et al 2016]. The rewards obtained by using quality metrics correspond with a positive float number.…”

Section: Formalizing the Model Repair Problem As A Markov Decision Prmentioning

confidence: 99%

“…We use Q-Learning as our learning algorithm and EMF as our modeling framework. As user preference, we decide to boost the maintainability of the models [Iovino et al 2020]. The maintainability quality metric considered in this paper has been defined according to the definition given in [Genero and Piattini 2001] which is based on some of the metrics shown in Table 1 as follows, According to the considered definition of maintainability the lower the values the better.…”

Section: Comparing Both Mdp Formalizationsmentioning

confidence: 99%

A comparative study of reinforcement learning techniques to repair models

Barriga

Mandow²,

Cruz³

et al. 2020

Proceedings of the 23rd ACM/IEEE International Conference on Model Driven Engineering Languages and Systems: Companion Proceedi

Self Cite

View full text Add to dashboard Cite

In model-driven software engineering, models are used in all phases of the development process. These models may get broken due to various editions during the modeling process. To repair broken models we have developed PARMOREL, an extensible framework that uses reinforcement learning techniques. So far, we have used our version of the Markov Decision Process (MDP) adapted to the model repair problem and the Q-learning algorithm. In this paper, we revisit our MDP definition, addressing its weaknesses, and proposing a new one. After comparing the results of both MDPs using Q-Learning to repair a sample model, we proceed to compare the performance of Q-Learning with other reinforcement learning algorithms using the new MDP. We compare Q-Learning with four algorithms: Q(λ), Monte Carlo, SARSA and SARSA (λ), and perform a comparative study by repairing a set of broken models. Our results indicate that the new MDP definition and the Q(λ) algorithm can repair with faster performance. CCS CONCEPTS • Software and its engineering → Model-driven software engineering; • Theory of computation → Reinforcement learning.

show abstract

Model Repair with Quality-Based Reinforcement Learning.

Cited by 14 publications

References 8 publications

Multi-Model Evolution through Model Repair.

Multi-Model Evolution through Model Repair.

ModelSet: a dataset for machine learning in model-driven engineering

A comparative study of reinforcement learning techniques to repair models

Contact Info

Product

Resources

About