$$\mathsf {Uhura}$$ : An Authoring Tool for Specifying Answer-Set Programs Using Controlled Natural Language

Abstract

In this paper, we present the tool \(\mathsf {Uhura}\) for developing answer-set programs by means of specifying problem descriptions in a controlled natural language which then are translated into answer-set programming (ASP) rules. The tool is aimed for supporting users not familiar with answer-set programming—or logic-based approaches in general—for developing programs. \(\mathsf {Uhura}\) is based on a new controlled natural language called \(\mathsf {L}^{\text {U}}\), which is in turn an adaption of \(\mathsf {PENG}^{\text {ASP}}\), a controlled natural language employed in the PENG ASP system, developed by Guy and Schwitter, for solving computational problems by translating \(\mathsf {PENG}^{\text {ASP}}\) statements into answer-set programs. In contrast to \(\mathsf {PENG}^{\text {ASP}}\), \(\mathsf {L}^{\text {U}}\) allows for a more natural translation into ASP rules and provides also a broader set of pre-defined sentence patterns. \(\mathsf {Uhura}\) is implemented in \(\mathsf {Java}\) and employs \(\mathsf {DLV}\) as backend answer-set solver.