Premium
Event tree algorithms to generate test sequences for composite Web services
Author(s) -
Endo Andre Takeshi,
Simao Adenilso
Publication year - 2019
Publication title -
software testing, verification and reliability
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.216
H-Index - 49
eISSN - 1099-1689
pISSN - 0960-0833
DOI - 10.1002/stvr.1637
Subject(s) - computer science , algorithm , event (particle physics) , tree (set theory) , test suite , web service , sequence (biology) , test case , greedy algorithm , graph , test (biology) , theoretical computer science , mathematics , combinatorics , machine learning , programming language , physics , regression analysis , quantum mechanics , biology , genetics , paleontology
Summary In service‐oriented architecture testing, the behaviour of a composite service and its partner services can be represented as an event graph (model). As test cases are derived from such a model, the Chinese postman problem (CPP) algorithm is applied to cover all edges. Although CPP provides an optimal solution, there is a lack of alternative algorithms to obtain not only a faster execution but also a satisfactory solution. This paper introduces the event tree algorithm (ETA) to generate test cases from event sequence models for Web service compositions. Two strategies are also proposed to improve ETA: First, a greedy strategy to build the tree is defined based on the cost of expanding given edges. Second, redundant elements are eliminated (pruned) in the tree. Four versions of ETA are compared with CPP with respect to test suite size, generation time, number of test sequences, and test sequence length.