Methods and Methodologies for Developing Answer-Set Programs

Answer-set programming (ASP) is an important formalism for declarative problem solving and nonmonotonic reasoning, having a model-oriented view for knowledge representation, i.e., models constitute answers to problems rather than proofs as in traditional logic-based approaches. Although different realisations of the ASP paradigm exist, it is usually identified with logic programming under the answer-set semantics, which we shall deal with in this project. The high relevance of this particular formalism lies not only in its intuitive character but is also due to the availability of sophisticated solvers, which lead to successful applications in diverse areas like diagnosis, planning, database repair, and Semantic-Web reasoning. However, a main obstacle for a wider acceptance of ASP is a lack of techniques for supporting the development of answer-set programs. Indeed, the need for tools, methods, and methodologies that ease the programming process has been clearly recognised in the ASP community, but, except for some preliminary work, no systematic study in this direction has been put forth so far. In this project, we address this challenging issue by providing adequate principles and techniques that are aimed to support the answer-set programmer during development. Generally, the focus of our research will be on methodologies for systematic program development, program testing, and debugging. In particular, in working on these areas, special emphasis will be given to modular programming concepts and to the ability of the developed techniques to respect the declarative nature of ASP. To support a sufficient level of applicability, we plan to offer solutions not only for the core language of ASP but also for important extensions thereof that are commonly used and realised in various answer-set solvers. Moreover, we will consider complexity and decidability issues for the identified methods whenever relevant and provide restricted techniques and approximations where needed. The methods resulting from the project should be incorporated in an integrated development environment for ASP that combines straightforward as well as advanced techniques realising a convenient tool for developing answer-set programs. In order to reach our goals, we will mainly rely on formal, logic-based methods, suitable for innovative novel support techniques for a declarative language. The availability of intelligent development methodologies and tools will have a major impact on the field of ASP as a whole, significantly advancing the state of the art. Furthermore, with such techniques at hand, both expert as well as novice programmers will have an enhanced access to powerful declarative problem-solving machineries.

Answer-set programming (ASP) is a logic-based approach for declarative problem solving with notable applications in diverse areas like planning, diagnosis, configuration, systems biology, semantic-web reasoning, and many others. So far, research in ASP focused on the analysis of theoretical aspects, the deployment of applications, and the development of practicably efficient ASP solver technology. In this project, the important complementary aspect of studying methods for developing answer-set programs was addressed. Such methods are important for a further awareness of ASP outside academia.The main results of the project comprise, on the one hand, theoretical results concerning novel methods for testing, validation, verication, and debugging of answer-set programs, as well as studying methods for systematic program development, and, on the other hand, realising these methods in terms of an integrated development environment (IDE) for ASP called SeaLion in order to make them available for practical applications.