Exporting Natural Language: Generating NL Sentences Out of S-BPM Process Models

Abstract

Assuming that a formal theory such as S-BPM is serving as a source for generation, it is possible to map certain model constructs on to a subset of the natural language and to formulate sentences with them. Hence natural language is a possible output of the S-BPM method. Due to the close relation between the syntax of the S-BPM method and the syntax of a natural language, this mapping can also be rather simple. This paper identifies at least four potential users within a business process and their requirements: Human actors who are assigned a role in the business process, people responsible for negotiating the service level agreements, requirement analysts and test analysts. Then the basic S-BPM elements are introduced as well as some modeling guide lines that ensure proper semantics of the S-BPM model and for the exported natural language. The required functionality and the definition of NL sentences and their elements are shown as well as the mapping and the algorithms needed to generate NL output.