Natural Language Generation for Legal Expert System and Visualization of Generation Process

An HPSG-based grammar and a sentence generation system for small set of Japanese in legal expert domain are constructed. The system adopts its own general semantic system into which a domainspeci c logical form is converted. This separation between domain-speci c semantics and linguistic one gives exibility to both the task processing and the sentence generation. We also propose a visualization system which shows the generation process in a tabular form and operates as a graphical user interface for grammar debugging.