Bidirectional Use of Knowledge in the Multi-modal NL Access System XTRA

The acceptability and effectiveness of an expert system is critically dependent on its user interface. Natural language could be a well-suited communicative medium; however, current NL interfaces to expert systems show an impeding number of shortcomings. In this paper, it is shown how the XTRA system a German NL interface between a user and one of various expert systems employs a framework of novel AI techniques to overcome these shortcomings: it uses various highly interacting knowledge sources for domain-independent linguistic knowledge and for domain-specific world knowledge. This supports the view of NL as a universal communicative medium. In order to allow for a communicatively adequate dialog processing, knowledge sources are bidircctionaly used by both the analysis and the generation component. Both are enhanced with knowledge about the visual context of the dialog situation and about the capability to perform pointing gestures. SB-ONE a knowledge representation formalism especially tailored for NL processing deals with these types of knowledge. Together with some important tools of this formalism it constitutes the core of the system, whose performance is demonstrated with the help of an example dialog. 1 Theoretical Foundations of XTRA Since expert systems and other large knowledge based systems are frequently employed in (prototypical) practical service, it turns out that the acceptability and, therefore, the effectiveness of an expert system (XPS) depends critically on its user interface ([Wahlster 86a,b, Hendrix 86, Tennant 861). Surely natural language would be a well-suited communicative medium if sufficiently mastered by the system. The XTRA system 1 is being developed in a basic research project which investigates open questions in the field of natural language processing, guided by the application as a (German) natural language (NL) interface to various expert systems . Current NL interfaces show an impeding number of shortcomings. They are often tailored to the underlying system with respect to different aspects. One aspect is the direct translation of the input into the XPS specific knowledge representation ([Datskovsky Moerdler et al. 871), and vice versa in the case of generation. As a consequence, linguistic and domain-specific processing are intertwined. When the domain is changed, even linguistic processing has to be built up from the scratch ([Cohen and Jones 891). Another problem in NL processing is the lack of knowledge representation formalisms common and sufficient for all internal processes. One aim of the XTRA project is to overcome the mentioned drawbacks in NL interfaces to at least some extent. Special emphasis is laid upon the separation of linguistic and world knowledge, the bidirectionality of knowledge sources, and the combination of pointing gestures and written NL. 1.1 Linguistic and world knowledge Knowledge about language is basically invariant, i.e. independent of a specific domain and, therefore, universal in any communicative situation. XTRA is equipped with two distinct (however: highly interacting) knowledge sources, the Functional-Semantic Structure (FSS) and the Conceptual Knowledge Base (CKB). The FSS's criteria for entailments and structuring are of linguistic nature, whereas the criterion of the CKB is its suitability for processing, e.g. drawing inferences and interacting with the XPS. As [Wilensky 87] noted, it is rather pointless to claim that a sentence's representation based on pure linguistic knowledge is equivalent to its meaning. But it is beyond doubt (as is agreed by e.g. [Jacobs 85] and even [Wilensky 87]) that it has important merits in NL processing. We would by no means say that understanding or generating language is free from extra-linguistic influences; but it is necessary to explicitly represent the linguistic properties as well. In XTRA, the FSS is seen as the intermediate structure during analysis or generation of an utterance. Semantical well-formedness according to a Fillmore-likc deep case classification ordered in a taxonomic hierarchy and encoded in the knowledge representation language SB-ONE (see Fig. 1) is requested for any in/output. The construction of the CKB (sec Fig.2) should be independent of linguistic considerations to reflect the relations necessary for adequately modelling the domain of the underlying XPS. Figure 1: The taxonomy of the Functional-Semantic Structure (FSS)