Automated Test Production -- Systematic Literature Mapping
José Marcos Gomes, Luis Alberto Vieira Dias
TL;DR
The paper surveys Automated Test Production (ATP) by conducting a systematic literature mapping to identify the current state of the art, open questions, and progress in ATP, particularly for practitioners developing internal enterprise software and augmenting generic applications. It follows Petersen's 2008 methodology, with planning, conduction, and analysis steps, using a targeted search of major publication databases for literature from 2015–2020 and classifying studies via dual facets (research type and contribution type). Key findings show a sustained level of ATP research across conferences and journals, a predominance of practical artifacts (tools, models, metrics) and evidence of code/data generation across studies, and no single venue dominating the field. The work lays groundwork for a future, more detailed investigation into generative testing within ATP and the evaluation of proposed solutions, providing a baseline for researchers and practitioners to track progress and identify opportunities in enterprise software testing.
Abstract
The broader goal of this research, on the one hand, is to obtain the State of the Art in Automated Test Production (ATP), to find the open questions and related problems and to track the progress of researchers in the field, and on the other hand is to list and categorize the methods, techniques and tools of ATP that meet the needs of practitioners who produce computerized business applications for internal use in their corporations - eventually it can be extended to the needs of practitioners in companies that specialize in producing computer applications for generic use.