Large-scale natural language generation requires the integration of vast amounts of knowledge: lexical, grammatical, and conceptual.

This paper describes a practical method of automatic simultaneous interpretation utilizing an example-based incremental transfer mechanism. We primarily show how incremental translation is achieved in the context of an example-based framework. We then examine the type of translation examples required for a simultaneous interpretation to create naturally communicative dialogs. Finally, we propose a scheme for automatic simultaneous interpretation exploiting this example-based incremental translation mechanism. Preliminary experimentation analyzing the performance of our example-based incremental translation mechanism leads us to believe that the proposed scheme can be utilized to achieve a practical simultaneous interpretation system.

This paper discusses a sentence generation system protector which uses: (i) a non-hierarchical semantic representation which allows for flexible lexical choice and uniform treatment of different languages, (ii) a lexicalised D-Tree Grammar which is very similar to Tree-Adjoining Grammar in spirit, and (iii) dynamic programming techniques to avoid doing redundant computations. We review the motivation for choosing such an organisation of the generation system and give an example of the generation of a sentence which involves a lexical gap. The generation of the example sentence requires a nondeterministic mode of computation (the lexical gap forcing backtracking). We show how dynamic programming techniques can be used to save re-generating structures using a top-down generation algorithm.