A method of functional testing for software components according to model-based testing techniques is proposed. Test cases are generated from UML diagrams and OCL constraints that comprise a component interface and realization specification. The method uses a reduced set of UML artifacts that constitute the main requirements for its application along with a component development process, making use of development artifacts. Also, the set of generated test artifacts are packed together with the provided components to reduce the overall testing effort when clients assemble applications. A tool has been developed to automate the method with test cases generated as Java test components. Test execution and result analysis is also supported. For each component, the tool generates a test component that can be easily upgraded and configured for testing the services provided by a component throughout its life cycle.

No abstract

No abstract

Automatic generation of built-in contract testers that check pairwise interactions between client and server components is presented. Test artifacts are added to development models, i.e., these models are refined towards testing. The refinement is specified by a set of ATL transformation rules that are automatically performed by the ATL-DT tool. We focus on the KobrA development methodology and the Built-In Testing (BIT) method. Therefore, development and testing artifacts are specified by UML diagrams. To make transformation from development to testing models possible, a UML 2 profile for the BIT method is proposed. The overall solution is part of the MoBIT (Model-driven Built-In contract Testers) tool that was fully developed following Model-Driven Engineering (MDE) approaches. A case study is presented to illustrate the problems and solutions addressed.

Example-dependent cost-sensitive classification methods are suitable to many real-world classification problems, where the costs, due to misclassification, vary among every example of a dataset. Tax administration applications are included in this segment of problems, since they deal with different values involved in the tax payments. To help matters, this work presents an experimental evaluation which aims to verify whether cost-sensitive learning algorithms are more cost-effective on average than traditional ones. This task is accomplished in a tax administration application domain, what implies the need of a cost-matrix regarding debt values. The obtained results show that cost-sensitive methods avoid situations like erroneously granting a request with a debt involving millions of reals. Considering the savings score, the cost-sensitive classification methods achieved higher results than their traditional method versions.