The Role of Similarity in Detecting Feature Interaction in Software Product Lines

Khoshmanesh, Seyedehzahra; Lutz, Robyn R.

doi:10.1109/issrew.2018.00020

Cited by 8 publications

(7 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Results from application and evaluation on a small product line showed that the best ML algorithms achieved good accuracy (0.75 to 1) for classifying productline feature interactions as unwanted or wanted. Directions for future work are to evaluate our link-prediction method on a larger product line and to investigate whether incorporating structural similarity measures, as in [19], improves classification of feature interactions.…”

Section: Discussionmentioning

confidence: 99%

“…Fig. 2 shows the intuition behind how using similarity indexes between two nodes in an interaction graph helps to detect new or missing unwanted feature interactions in a new version or product-line product [18][19][20][21]. As shown in Fig.…”

Section: Overviewmentioning

confidence: 99%

“…This is because similar features have been observed to behave in similar ways. If there is a feature in the feature interaction graph that contributes to some unwanted feature interactions, the features which are similar to this feature often will contribute to the same unwanted interactions [19][20][21].…”

mentioning

confidence: 99%

“…The work described in this paper is part of our ongoing effort toward improved detection of feature interactions during the requirements analysis of a new product. In earlier work we studied similarity measures based on features' structural elements (classes, attributes and methods) [19] and on features' relative positions in a software product line's feature model [19][20][21]. New work that is first reported in this paper is our representation of the feature interaction problem as a link prediction problem, which makes feature interactions amenable to classification (wanted/unwanted) and the building of a predictive learning model, together with the evaluation results from our initial application of this approach.…”

mentioning

confidence: 99%

See 3 more Smart Citations

Does Link Prediction Help Detect Feature Interactions in Software Product Lines (SPLs)?

Khoshmanesh,

Lutz

2020

Preprint

Self Cite

View full text Add to dashboard Cite

An ongoing challenge for the requirements engineering of software product lines is to predict whether a new combination of features (units of functionality) will create an unwanted or even hazardous feature interaction. We thus seek to improve and automate the prediction of unwanted feature interactions early in development. In this paper we show how the detection of unwanted feature interactions in a software product line can be effectively represented as a link prediction problem. Link prediction uses machine learning algorithms and similarity scores among a graph's nodes to identify likely new edges. We here model the software product line features as nodes and the unwanted interactions among the features as edges. We investigate six link-based similarity metrics, some using local and some using global knowledge of the graph, for use in this context. We evaluate our approach on a software product line benchmark in the literature, building six machine-learning models from the graph-based similarity data. Results show that the best ML algorithms achieved accuracy of 0.75 to 1 for classifying feature interactions as unwanted or wanted in this small study, and that global similarity metrics performed better than local similarity metrics. The work shows how link-prediction models can help find missing edges, which represent unwanted feature interactions that are undocumented or unrecognized, earlier in development.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Overviewmentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

See 2 more Smart Citations

Does Link Prediction Help Detect Feature Interactions in Software Product Lines (SPLs)?

Khoshmanesh,

Lutz

2020

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…• A feature interaction occurs when we integrate two or more resources to produce a new product, but together they do not work as intended [24]; • A feature interaction is the way a resource or a set of resources modifies or influences another resource in its behavior in the system [30]; • A feature interaction is a situation in which two or more characteristics exhibit unexpected behavior that does not occur when the characteristics are used in isolation [5]; • A feature interaction occurs when a combination of specific characteristics has an unexpected influence on performance [52]; and • A feature interaction is the combination of two or more services that perform correctly and individually to obtain unexpected results when combined [31]. As far as there is still no consensus regarding the definition of a feature interaction, we redefine it from the point of view of smart environments.…”

Section: Formal Definitionsmentioning

confidence: 99%

Feature Interaction in Smart Environments

Jesús¹,

Claro²,

Rios³

2021

Preprint

View full text Add to dashboard Cite

The Internet of Things (IoT) connects many devices daily together in the same environment. Each device may follow the set of rules from a static environment. A static environment is usually controlled by an expert who knows all the necessary rules to provide this environment. The violation of one rule can cause a feature interaction. A feature interaction occurs when two or more devices generate instability in an environment. In a dynamic environment like IoT, devices' inclusion, and exclusion make it impossible for an expert to maintain all these rules up-to-date. It is necessary to provide an automatic solution to avoid violating these rules and maintain the environment's good performance. Thus, this work introduces a new approach to detect a feature interaction in dynamic environments automatically. Almost all previous work provide static rules defined by an expert in a controlled environment to detect an interaction. However, this is not possible in dynamic environments because of the number of device interactions and the number of device connections in/out, which grow exponentially in IoT environments. We started with a lightweight systematic review to better position our research, and then we identified one gap to provide our solution. Thus, our method learns to detect the interactions based on data analysis and then automatically predict the device detections in IoT environments. Datasets were manually annotated. Experiments were performed, and results provide evidence that automatic detection of a set of device interactions is possible in similar or either in complementary domains.

show abstract

Leveraging Feature Similarity for Earlier Detection of Unwanted Feature Interactions in Evolving Software Product Lines

Khoshmanesh

Lutz

2019

Similarity Search and Applications

View full text Add to dashboard Cite

The Role of Similarity in Detecting Feature Interaction in Software Product Lines

Cited by 8 publications

References 41 publications

Does Link Prediction Help Detect Feature Interactions in Software Product Lines (SPLs)?

Does Link Prediction Help Detect Feature Interactions in Software Product Lines (SPLs)?

Feature Interaction in Smart Environments

Leveraging Feature Similarity for Earlier Detection of Unwanted Feature Interactions in Evolving Software Product Lines

Contact Info

Product

Resources

About