Knowledge Representation with MESNET : A Multilayered Extended Semantic Network
Semantic Networks (SN) have been used in many applications, especially in the field of natural language understanding (NLU). The multilayered extended semantic netword MESNET presented in this paper on the one hand follows the tradition of SN starting with the work of Quillian . On the other hand, MESNET for the first time consequently and explicitly makes use of a multilayered structuring of a SN built opon an orthogonal system of dimensions and especially upon the distinction between an intensional and a preextensional layer. Furthermore, MESNET is based on a comprehensive system of classificatory means (sorts an features) as well as on semantically primitive relations and functions. It uses a relatively large but fixed inventory of representational means, encapsulation of concepts and a distinction between immanent and situative knowledge. The whole complex of representational means is independent of special application domains. At the same time, it is fine grained enough to allow for the differentiation of all important nuances of meaning in the knowledge representation. MESNET has been especially developed for natural language understanding in question answering systems (QAS). A first prototype is successfully used for the meaning representation of natural languagge expressions in the system LINAS. In this paper, MESNET is presented in its double function as a cognitive model and as the target language for the semantic interpretation processes in NLU systems.
Nutzung und Vervielfältigung:
Alle Rechte vorbehalten