Text Structuring in a Multilingual System for Generation of Instructions

Abstract

The approach presented in this paper has been developed in the context of the international project called AGILE (Automatic Generation of Instructions in Languages of Eastern Europe)¹. The overall aim of the project is to develop a multilingual system for generating continuous instructional texts in Bulgarian, Czech and Russian [2]. The project is concerned mainly with (i) the development and adaptation of linguistic resources for the chosen languages, and (ii) the investigation and specification of text structuring strategies employed in those languages for the given type of texts. In the current paper, we concentrate on the latter issues.

What we are particularly interested in is the following. In the application scenario of the system developed in AGILE, the user provides a definition of a text content in the form of an A(assertion)-box. from this A-box we are then to generate a text, in either of the languages we are concerned with. The problem we address in this paper is how to go about this for relatively simple instructional texts about working with AutoCAD.

To that end, we begin with a brief discussion of the texts (and their language-dependent styles) we would like to be able to generate (Section 2). As we already pointed out, these texts are generated from (or “realize”) a text content specified by an A-box. In Section 3 we show how we can use the concepts, employed in an A-box, to set up a structure for a text that would realize the content given by that A-box. Finally, we discuss in Section 4 how the approach outlined in the previous sections is implemented in the KPML multilingual generation environment [1] that we use in the context of AGILE.

AGILE (Automatic Generation of Instructions in Languages of Eastern Europe) is an international project supported by the European Commission within the COPERNICUS programme, grant No. PL961104. We would like to thank our colleagues from Charles University, the University of Brighton, the University of the Saarland, the Bulgarian Academy of Sciences and the Russian Research Institute of Artificial Intelligence for their cooperation.