Abstract
This report discusses several approaches to Controlled English and the problem of verbal representations of rules and suggests a method for verbalizing the visual and formal representations of vocabularies and rules obtained by UML class models and OWL ontologies. The method, called I1 Controlled English (I1CE), is based on defining verbalization patterns for all important kinds of fact types (predicates), and combining them with the help of keywords expressing quantifiers, logical connectives and other sentence construction operators. An important issue is the quality of the verbalization results. It is discussed that the simple mapping techniques based on fact type verbalization patterns lead to results that are acceptable in many important cases but which could be further improved by applying more advanced techniques such as the DRS-based treatment of anaphora employed in the Attempto system.

URL:
http://rewerse.net/publications/rewerse-publications.html#REWERSE-DEL-2005-I1-D2

BibTeX:

@deliverable{REWERSE-DEL-2005-I1-D2,
	url = {http://rewerse.net/publications/rewerse-publications.html#REWERSE-DEL-2005-I1-D2}
}