Test Case Generation from Cause-Effect Graph Based on Model Transformation

Abstract

In software testing, cause-effect graph assures coverage criteria of 100% functional requirements with minimum test case. The existing test case generation from causeeffect graph implements the algorithmic approach. It has disadvantages to modify the entire program if the input model is different. In contrast, model transformation approach can flexibly implement with even a different input models. In the future, we need to study the method of automatic generation of test cases from UML Diagram. It is possible to generate the test case when mapping between cause-effect graph and UML diagram. In this paper, as a first research step, we propose the method to generate test cases from cause-effect graph based on model transformation. To implement the proposed method, we write the rules of model transformation with ATLAS Transformation Language (ATL), and execute the rules in development environment of Eclipse. The implemented tool of the proposed method can be easily extended by rewriting with the mapping rule between cause-effect graph and UML diagram. We just define the relationship between each models to generate the test case.

Forward-Propagation Approach for Generating Feasible and Minimum Test Case Suites From Cause-Effect Graph Specifications

“…Many approaches such as [36] and [37] relate cause-effect graphs to Unified Modeling Language (UML) models and offer ways of automatic conversion between the two representations in order to convert object-oriented programming domain problems into the cause-effect graphing domain, which is more suitable for black-box testing. The proposed algorithms use Object Constraint Language (OCL) expressions or the ATLAS programming language for creating cause-effect graph representations and generating test case suites from UML system models.…”

“…Graphs defined in [3] will be used for validating algorithm performances on large graphs. The following original algorithm enhancements and applications of the initial algorithm will be used for comparison: binary tree modification from [34], college placement process application from [41], model transformation from [37], mathematical formalization from [4], software tool implementation from [12], new minimization approach from [8] and ATP speed train application by [10].…”

Generating Test Cases from Role-Based Access Control Policies using Cause-Effect Graph

“…Chung [14] developed a fault model for the cause-effect graph. Son, et al, [15] proposed a method to generate test cases based on model transformation from cause-effect graph and UML Diagram.…”

CEREC: Causality Extraction from Requirements Artifacts

“…Extracting this relationship is, for example, used in the medical domain to gather causal relations between symptoms and diseases and deriving insights for diagnostics [2], for question answering [3], and in the economic domain, where causal relations can be used to unravel insight on stock market movements [4]. The domain of requirements engineering is eligible for causality extraction for three reasons: (1) requirements are predominantly written in natural language [5], (2) formalized requirements are reusable for other artifacts like test cases [6], and (3) causal relations in requirements convey important conditions, which explicitly define the described system. Existing approaches to causality extraction fail to translate into requirements engineering for two reasons: existing approaches are (1) based on semantic causality, while relations described in requirements engineering are commonly not of semantic nature, and (2) they only extract single words as causal events, which omits important information necessary to reuse the extracted causality for further processing [7].…”