Premium
PLANNING TO FIND THE REFERENTS OF NOUN PHRASES
Author(s) -
Haas Andrew R.
Publication year - 1995
Publication title -
computational intelligence
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.353
H-Index - 52
eISSN - 1467-8640
pISSN - 0824-7935
DOI - 10.1111/j.1467-8640.1995.tb00051.x
Subject(s) - referent , noun phrase , utterance , computer science , phrase , linguistics , noun , artificial intelligence , natural language processing , natural (archaeology) , proper noun , table (database) , robot , natural language , philosophy , archaeology , history , data mining
Finding the referent of a definite noun phrase is a classic problem in natural language processing. Most work assumes that when a program begins to analyze an utterance, it already has the knowledge it needs for identifying referents. If a robot accepts commands in natural language, this assumption may not hold. Suppose a user says, “Go get the book on the table in room 3,” but the robot has never been in room 3, so it lacks the knowledge it needs to identify the referent of “the book on the table in room 3.” To gain this knowledge, the robot must travel to room 3 and search. This article describes a simulated robot that accepts commands in English and actively searches for the referents of definite descriptions.