Excerpt from Directions in Artificial Intelligence: Natural Language Processing

Most recently, a paper that hasn't gotten around very much, is a very ambitious effort by Roger Schank [9] on setting up discourse structure for paragraphs. That paper will probably be circulating pretty widely by Spring. In that, he very ambitiously takes the folk tale from Eskimo literature several paragraph stories about 5 I think and he makes a causal chain of the relationships among the deep conceptual dependency structures that he derives. It is easily criticizable. There are many things where each of us would differ on whether that is the path or exactly what is going on. But I admire it very much because it is a very ambitious attempt to deal with a fairly large piece of text. The point is we don't have models for dealing with large amounts of text, or even with several paragraphs of text. We don't really know how to handle the semantics of text discourse.

The basic problems are: How do you use sentences to understand following sentences? How do you solve problems of reference? And what is an adequate lexical structure to do this? What is an adequate semantic structure to represent the resulting meanings? That whole family of questions needs to be answered.

We are very much concerned in my group at Texas with studying all these things. We have one student (we don't have a big budget for this) but one student is working on discourse analysis a dissertation by Hendrix [10] is about finished on modelling techniques an approach toward representing meanings as set theoretic expressions. It takes semantic nets right down to the abstract algebraic description of what the meaning of sentences is. And what I will talk about today is Robert Amsler's studies of the Merriam - Webster dictionaries.

About the Publisher

Forgotten Books publishes hundreds of thousands of rare and classic books. Find more at www.forgottenbooks.com

This book is a reproduction of an important historical work. Forgotten Books uses state-of-the-art technology to digitally reconstruct the work, preserving the original format whilst repairing imperfections present in the aged copy. In rare cases, an imperfection in the original, such as a blemish or missing page, may be replicated in our edition. We do, however, repair the vast majority of imperfections successfully; any imperfections that remain are intentionally left to preserve the state of such historical works.