Requirements-based testing with defect taxonomies

Abstract

In this paper we summarize requirements-based testing with defect taxonomies which seamlessly integrates defect taxonomies into the standard test process to improve the effectiveness and the efficiency of testing requirements. Defect taxonomies which consist of hierarchies of defect categories provide information about the distribution of faults and failures in a project. In practice, most defect taxonomies are only used for the a-posteriori allocation of testing resources to prioritize failures for debugging purposes. Requirements-based testing with defect taxonomies (RTDT), which has been defined and evaluated in several previous publications [FB12], [FB13a], [FB13b], [FB14a], [FB14c], [FBP14], exploits the full potential of defect taxonomies to control and improve all phases of the overall test process, i.e., test planning, design, execution and evaluation. Figure 1 summarizes the process steps and artifacts of RTDT. 2: Link and Validate Requirements and Defect Taxonomy Defect Taxonomy Requirements Specification Defect Category of Beizer DC Description of DC Severity 1: Create Defect $\cdots \cdots \cdots $REQ Description Priority Taxonomy $\cdots \cdots 15$ Data storage high 4xxx . Data D1 Incorrect access $\cdots $normal 27 Search function medium 42xx . . Data access . D2 Erroneous save of critical data critical 40 Documentation high $\cdots \cdots \cdots \cdots \cdots \cdots \cdots \cdots 3$: Plan Testing Test Strategy ID Test Design Test Strength 1 Test Strength 2 Test Strength 3 Technique (low) (normal) (high) TDS: Priority Severity Test Strength TDS1 Process cycle Main paths Branch coverage Loop coverage 6: Evaluate Sequence tests hi gh bl ocker, cri ti ca l , ma jor 3 oriented $\cdots \cdots \cdots \cdots \cdots $Tests norma l bl ocker, cri ti ca l $3 \cdots \cdots \cdots \cdots \cdots $TDD: norma l ma jor, norma l , mi nor, tri vi a l 2 Data TDD2 EP: Equivalence EP valid EP valid+invalid EP valid+invalid Partitioning l ow mi nor, tri vi a l