Domenico Parisi

Disambiguation in a Lexically Based Sentence Understanding System

Publisher Summary This chapter discusses a number of problems of disambiguation within a lexically-based model of sentence understanding system. Syntactic knowledge contributes to disambiguation through the disambiguation procedures. These procedures are carried by the lexical entries and applied to the preceding context and through the four well-formed principles that signal failures in syntactic garden path sentences. Encyclopedic knowledge operates through spreading activation and shortest path finding to select contextually appropriate readings of ambiguous words and to solve a number of additional problems of deep understanding. Currently the system is being revised to generate a more connectionist implementation with syntactic and encyclopedic knowledge represented in a more uniform way, in which understanding is obtained by an interaction of the two types of knowledge in a shared network. In the new implementation, disambiguation is brought about by following alternative paths in a parallel way, and waiting for both preceding and subsequent contexts, to send additional activation to one alternative and eliminate the others.

GEMS: a model of sentence production

The paper describes GEMS, a system for Generating and Expressing the Meaning of Sentences, focussing on the generation task, i.e. how GEMS extracts a set of propositional units from a knowledge store that can be expressed with a well-formed sentence in a target language. GEMS is lexically distributed. After a central processor has selected the first unit(s) from the knowledge store and activated the corresponding lexical entry, the further construction of the sentences meaning is entrusted to the entries in the vocabulary. Examples of how GEMS constructs the meaning of a number of English sentence types are briefly described.