Historical Perspective

• During 1970s AI Systems were developed that demonstrated interesting aspects of language understanding in restricted domains. (e.g. Winograd (1972), Woods (1977), Waltz (1978)).
• During 1980s continued progress on developing NLP systems using hand-coded grammars and knowledge bases (Allen 1987)
• All such systems required a great deal of domain-specific knowledge engineering and as a result were quite brittle.
• In recent years, focus has shifted from rationalist methods to empirical or corpus-based methods in which development is data driven.