tailieunhanh - Báo cáo khoa học: "Panel on Natural Language and Databases"

While I disagree with t h e proposition t h a t d a t a b a s e query has outlived its usefulness as a test e n v i r o n m e n t for n a t u r a l language processing (for reasons t h a t I give below), I believe t h e r e are o t h e r reasonable tasks which can also s p u r new research in NL processing. In particular, I will suggest t h a t the task of providing a. | Panel on Natural Language and Databases Daniel p. Flickinger Computer Research Center Hewlett-Packard Company 1501 Page Mill Road Palo Alto California 94304 USA While I disagree with the proposition that database query has outlived its usefulness as a test environment for natural language processing for reasons that I give below I believe there are other reasonable tasks which can also spur new research in NL processing. In particular I will suggest that the task of providing a natural language interface to a rich programming environment offers a convenient yet challenging extension of work already being done with database query. First I recite some of the merits of continuing research on natural language within the confines of constructing an interface for ordinary databases. One advantage is that the speed of processing is not of overwhelming importance in this application since one who requests information from a database can expect the retrieval to take time with or without a natural language interface. Of course speed is desirable and waiting for answers to apparently simple requests will be irritating but some delay will be tolerable. This tolerance on the part of the user will I suggest disappear in applications where an attempt is made to engage a system in dialogue with the user cis would be the case in some expert systems or in teaching systems. Assuming that natural language systems will not by themselves get faster as they are made to cope with larger fragments of a natural language it will be useful to continue with database query while we wait for miracles of technology to fill our demands for ever greater processing speed. A second reason for not yet abandoning the database query as a test environment is that a great deal of important natural language processing research remains to be done in generalizing systems to cope with more than one natural language. Work on language universals gives reason to believe that some significant part of a natural .