Providing natural language assistance in locating objects: a general model for information selection and generation
Ronnie W. Smith · 1988
This paper presents a computational method for the selection and generation of information to be used in providing natural language assistance in locating objects.This method's significance is that it integrates physical information about the situation, previous discourse history, and the level of user experience in producing its responses.Background information from cognitive psychology that provides the ideas for this method is presented.Also discussed is GALOP, a program written to demonstrate the validity of this method.Results from tests on the appropriate verbosity level for responses by GALOP for different classes of users axe also presented.To the left of spring 17 is spring 16. (where spring 17 has recently been referenced) or