It creates some plans and tries to execute them. It analyses the situations deeper in the tree only if the plan fails. In that case it generates new plans correcting what is wrong in the old one. So, the program considers only natural branches of the tree. It can find combinations for which it is necessary to look more than twenty ply ahead. The paper describes the methods used for analyzing a situation and for modifying unsuccessful plans. Then we examine some results found by the program.

Rote learning.We can keep all the situations already found. With each situation we store an indication on its interest or the move which has to be played. Samuell gives an example of such an application. This can be done if there are not too many possible situations. Even in games where there are many possible situations, this method can be useful for the beginning or the end of the games. We can improve this method: if the rulesare the same for all the players, we can standardize the situations: we assume that it is always the same player who has to play; for instance, at chess, white. We just keep half of the possible situations. At Go Moku where there are two axes of symetry, we just keep a quarter of the situations. But even with these improvements, there are many cases where this method is not useful because there are too many situations. It is doubtful that we can have good results in the middle game 398 at chess with such methods. We can try to generalize what has been done in a situation to another similar situation. For example in the second situation we play the same move than in the first one. Waterman2 has written an interesting progrpi playing poker. Let us describe it roughly. A situation is described by the value of seven variables: value of the program's hand, amount of money in the pot, measure of conservative style by the opponent... The program defines a partition of the set of possible values of these variables. For instance: If our hand is excellent, bet low if the opponent tends to be a conservative player and has just bet low. The problem is to define wisely these subsets. This can be done by the program which improves progressively the quality of the partition. This method is good for poker and it obtained very good results. But it is difficult to see how we can use it in a game like chess. How could we evaluate the similarity between situations, such as in similar situations we have to play the same move? A different position of a pawn can destroy a combination.

A. YONEZAWA and C. HEWITT 41 REPRESENTATIONS FOR ABSTRACT REASONING 4 A maximal method for set variables in automatic theorem-proving W. W. BLEDSOE

References are to text of chapters, not to the bodies of bibliographies.

Donald Michie Volumes 1 --7 are published by Edinburgh University Press and in the United States by Halsted Press (a subsidiary of John Wiley & Sons, Inc.) Volumes 8 -- 9 are published by Ellis Horwood Ltd., Publishers, Chichester and in the United States by Halsted Press (a subsidiary of John Wiley & Sons, Inc.) MACHINE INTELLIGENCE 9 New York - Chichester - Brisbane - Toronto First published in 1979 by ELLIS HORWOOD LIMITED Market Cross House, Cooper Street, Chichester, West Sussex, P019 lEB, England The publisher's colophon is reproduced from James Gillison's drawing of the ancient Market Cross, Chichester No part of this publication may be reproduced, stored in a retrieval system or transmitted, in any form of by any means, electronic, mechanical, photocopying, recording or otherwise, without prior permission. One intelligent approach to prefaces -- is to have the empty preface. The well prepared reader will form a good idea of the technical programme just from looking at the table of contents; together with the names of the authors, this gives him a good idea of what happened at the symposium. I could try to assess the tallcs and direct the reader's attention to the more interesting communications. But I fear this would be too subjective and unfair to the remaining authors -- all of them equally represented in this book. However, recalling that Spring week in Repino, a resort 20 kilometres from Leningrad on the Bay of Finland and unpopulated at that time of year, I have come to the definite conclusion that the scientific meeting was in its own way unique.

For the past ten years we have been working on the problem of getting a computer to understand natural language.

A model for adaptive problem solving applied to natural language acquisition.

J. Do RAN 433 22 Chromosome classification and segmentation as exercises in knowing what to expect.

The junction categories and link planting rules of SEE lyzed. That, however, is not the main point; it is merely typical of the way in which the program developed by a process of finding counter-examples that both invalidated old rules and hinted at new ones (Winston, 1973). The need to add and modify rules almost continuously to handle exceptions suggests that there is a basic flaw in the design. The flaw seems to be that Guzman used locally computed picture predicates as evidence for global scene-based properties. To avoid this one must ask what do the lines in the picture depict?

Many of the problems of recognition and interpretation encountered in archaeology have close parallels with classic artificial intelligence problems, notably those of scene analysis.