The program has two modes: PARAPHRASE and INFERENCE. In PARAPHRASE mode up to 150 semantic paraphrases can be generated from an input sentence by reading the conceptual representation underlying that sentence using different words and concept combinatione.

Syntax, Preference and Right Attachment Yorick Wilks, Xiuming Huang & Dan Fass Computing Research Laboratory New Mexico State University Las Cruces, NM, USA 88003 ABSTRACT The paper claims that the right attachment rules for phrases originally suggested by Frazier and Fodor are wrong, and that none of the subsequent patchings of the rules by syntactic methods have improved the situation. For each rule there are perfectly straightforward and indefinitely large classes of simple counterexamples. We then examine suggestions by Ford et a!., Schubert and Hirst which are quasi-semantic in nature and which we consider ingenious but unsatisfactory. We offer a straightforward solution within the framework of preference semantics, and argue that the principal issue is not the type and nature of information required to get appropriate phrase attachments, but the issue of where to store the information and with what processes to apply it. We present a prolog implementation of a best first algorithm covering the data and contrast it with closely related ones, all of which are based on the preferences of nouns and prepositions, as well as verbs.

Fourth International Joint Conference on Artificial Intelligence Advance Papers, pp. 26B-273

"With the objective of making into a science the art of verifying computer programs (debugging), the author addresses both practical and theoretical aspects of the process. A classic of sequential program verification, this volume has been translated into almost a dozen other languages and is much in demand among graduate and advanced undergraduate computer science students. Subjects include computability (with discussions of finite automata and Turing machines); predicate calculus (basic notions, natural deduction, and the resolution method); verification of programs (both flowchart and algol-like programs); flowchart schemas (basic notions, decision problems, formalization in predicate calculus, and translation programs); and the fixpoint theory of programs (functions and functionals, recursive programs, and verification programs). The treatment is self-contained, and each chapter concludes with bibliographic remarks, references, and problems." New York: McGraw-Hill, 1974.

Mathematical decision theory can be combined with heuristic techniques to attack Artificial Intelligence problems. As a first example, the problem of breaking an image into meaningful regions is considered. Bayesian decision theory is seen to provide a mechanism for including problem dependent (semantic) information in a general system. Some results are presented which make the computation feasible. A programming system based on these ideas and its application to road scenes is described.

"New directions in Artificial Intelligence research have led to the need for certain novel features to be embedded in programming languages. This paper gives an overview of the nature of these features, and their implementation in four principal families of AI languages: SAIL; PLANNER/CONNIVER; QLISP/INTERLISP; and POPLER/POP-2. The programming featurcs described include: new data types and accessing mechanisms for stored expressions; more flexible control structures, including multiple processes and backtracking; pattern matching to allow comparison of data item with a template, and extraction of labeled subexpressions; and deductive mechanisms which allow the programming system to carry out certain activities including modifying the data base and deciding which subroutines to run next using only constraints and guidelines set up by the programmer." ACM Computing Surveys, Vol. 6, No. 3, September 1974, pp. 155 174.

The problem of representation and handling of constraints is here considered, mainly for picture processing purposes. A systematic specification and utilization of the available constraints could significantly reduce the amount of search in picture recognition. On the other hand, formally stated constraints can be embedded in the syntactic productions of picture languages. Only binary constraints are treated here, but they are represented in full generality as binary relations. Constraints among more than two variables are then represented as networks of simultaneous binary relations.

This paper describes BUILD, a computer program which generates plans for building specified structures out of simple objects such as toy blocks. A powerful heuristic control structure enables BUILD to use a number of sophisticated construction techniques in its plans. Among these are the incorporation of pre-existing structure into the final design, pre-assembly of movable sub-structures on the table, and the use of extra blocks as temporary supports and counterweights in the course of the construction. BUILD does its planning in a modeled 3-space in which blocks of various shapes and sizes can be represented in any orientation and location. The modeling system can maintain several world models at once, and contains modules for displaying states, testing them for inter-object contact and collision, and for checking the stability of complex structures involving frictional forces.

This paper presents two new machine learning procedures used to arrive at “knowledgeable” static evaluators for checker board positions. The static evaluators are compared with each other, and with the linear polynomial used by Samuel [9], using two different numerical indices reflecting the extent to which they agree with the choices of checker experts in the course of tabulated book games. The new static evaluators are found to perform about equally well, despite the relative simplicity of the second; and they perform noticably better than the linear polynomial. An indication of the significance of the absolute values of these two numerical indices is provided by a discussion of a simple, purely heuristic, static evaluator, whose performance indices lie between those of the polynomial and those of the other two static evaluators.

British Medical Journal, 1, 376–380.