Building natural language generation systems
Building natural language generation systems
Hi-index | 0.01 |
This paper presents an architecture for generating online and written technical documentation. Action plans are generated from an underlying model of the system being documented, which may be post-edited to refine the output. An NLG system is built based on this architecture allowing the generation of documentation in Russian and English. A range of built-in tools enable editing of linguistic and domain knowledge at any point in document processing. This architecture is intended to help an author select the right balance between automatic and controlled generation for a specific application.