Prioritizing Coverage-Oriented Testing Process — An Adaptive-Learning-Based Approach and Case Study

Model-based test case prioritization using cluster analysis: a soft-computing approach

“…Soft computing methods can therefore be used for finding approximate solutions for real-world problems that contain inaccuracies and uncertainties. Additionally, soft computing techniques do not rely on the assumptions that are common to conventional statistical methods, such as the underlying statistical distribution of data, and thus they are useful in situations where little or no prior knowledge exist [22,23].…”

“…The importance of this method comes to light especially when the source code of the system is unavailable [21][22][23].…”

“…Here, the optimal number of clusters, say c , is determined in advance by using the cluster validity algorithm, as described in [57]. Those c clusters are then obtained by means of both ACL and FCM clustering algorithms [21][22][23].…”

“…all [13,[21][22][23][24]. The importance index, imp(x i ), of the i th event belonging to the k th cluster is in this respect defined as follows:…”

“…One of the benefits of the proposed approach is that it can be deployed in any stage of software development to reflect the preferences of the tester wherever he/she needs to select a subset among all possible/available test cases in favor of the total testing cost. Although the authors' previous works [13,[18][19][20][21][22][23] also defined similar attributes for events and assigned values to them in order to determine the degree of preference by means of cluster analysis [24], the approach presented in this article is different in that it classifies the events using two techniques adapted from soft computing, namely neural networks and fuzzy logic [22,23]. Details and the theoretical basis of the aforementioned clustering-based prioritization approaches are given in the authors' previous works [21][22][23].…”