Book
175
reviews
Altos, CA. Unfortunately, the book is rendered illegible by innumerable misprints such as mis-spellings, word omissions and duplications, morpheme omissions and duplications, strange punctuation and word division style, etc. I stopped reading on page 26 because I was unable to tolerate more than two distorting misprints in one sentence. This German edition is useless. It would be wise of the publisher to retract it and to do better next time. The topics dealt with are: Declarative knowledge, Inference, Resolution, Nonmonotonic reasoning, Induction, Reasoning with uncertainty, Knowledge and belief, Metaknowledge and metareasoning, State and change, Planning, and Intelligent-agent architecture. The edition is not reviewable (see above). Christian E. Dralle Free Univesity of Berlin, Germany
Antony Galton (ed.), Press, Harcourt Brace hardcover, L 24.00.
Temporal
Jovanovich
Logics
and
Publishers,
Their
London,
Applications
1987) XII
(Academic + 244 pp.,
This book has arisen from a conference on Temporal logic held at the University of Leeds It contains six comprehensive, most in 1986 and organized by the editor, Dr. Galton. interesting, and highest quality papers on the philosophical foundations, axiomatization, and programming of logics for reasoning about time, and it is the first one in the literature to present new technical work on the subject from the perspective of artificial intelligence. The papers are well organized in six chapters, are entirely self-contained and can be read independently of each other. Antony G&on in his introductory article on ‘Temporal logic and computer science’ (pp. l-52) provides a lucid overview on the subject. He examines the philosophical background in which temporal logic was first developed. He then considers the application of ideas from Lastly, he examines the further temporal logic to the field of AI and Congitive Science. development of temporal logic in software engineering. Chapter 2 presents Howard Barringer’s paper on ‘The use of temporal logic in the compositional specification of concurrent systems’ (pp. 53-90) dealing with the application of temporal logics in the formal specification of complex computing systems. The relevance of compositional proof theories, modularity, and abstractness are discussed. A basic technique for obtaining compositional program proof theories, using a temporal language as the assertion language, is demonstrated. Several specializations for various real parallel programming language frameworks are indicated. Finally, some problems in obtaining abstract semantic descriptions of programming languages in the temporal framework are discussed, together with solutions suggesting the use of a logic based on a linear dense time model. A framework for ‘Temporal logic programming’ (pp. 91-119) is offered by Roger Hale in chapter 3. He describes and uses Moszkowski’s Interval Temporal Logic (ITL) whose unit of consideration is the interval as opposed to the individual state. Besides the operational interpretation he gives ITL, the main body of his paper comprises some examples to illustrate how to apply this framework. The examples are concerned with the representation of real physical systems, ranging from animation to data transmission. Fariba Sadri’s contribution in chapter 4 entitled ‘Three recent approaches to temporal reasoning’ (pp. 121-168) deals with the formalization of time and events. She describes and discusses Kowalski and Sergot’s calculus of events; Lee, Coelho, and Cotta’s logic of time and events; and Allen’s temporal logic. Finally, Hanks and McDermott’s recent work on temporal and default reasoning is briefly discussed, with a suggestion that it merits further investigation in the light of recent results on the semantics of negation as failure.
176
Book reviews
The editor presents a pioneer work on ‘The logic of occurrence’ (pp. 169-196) in chapter 5 where he discusses the formalization of event sentences within the modal framework exempAn axiomatization is given of three tense operators lified by A.N. Prior and his followers. for past, present, and future. Corresponding proof theories are discussed, and their soundness and completeness are proved. It is also shown how the formalism might be adapted to encompass explicit temporal reference. In the final chapter 6, a second pioneer work on ‘Modal and temporal logic programming’ and (pp. 197-237) is p resented by Dov Gabbay. He extends Horn logic to allow for temporal alethic-modal connectives. The computation procedures are given and soundness is proved. It is shown that computing in modal and temporal logic is like computing in classical logic without Skolemizing. The volume is indispensable to the library of AIM researchers concerned with the problems It is of excellent printing quality, of temporal reasoning in designing medical AI systems. and compared to the unjustifiably high prices of German books, its pricing is a gift.
Humboldt
University,
Jim Hunter, John Cookson, and Jeremy Wyatt (eds.), Verlag, Berlin, 1989) X + 330 pp., softcover, DM 81.00.
Dagmar Berlin,
AIME
GrSppel Germany
89 (Springer-
This bare volume, void of any introduction and index, presents the proceedings of the Second European Conference on Artificial Intelligence in Medicine held on August 29-31, 1989 in London. It offers 36 papers organized in 10 chapters on knowledge elicitation, architectures for knowledge-based systems, methodology, and clinical applications. The topics dealt with range from clinical judgment analysis to clinical expert systems and uncertainty management. The papers differ considerably in their level of sophistication and the information they provide. Especially the theoretical ones concerned with the methodology of AIM research are of dubious value and shed some light on the immaturity of AIM foundations. It seems that most researchers in this area are not well acquainted with the meaning and appropriate use of basic methodological concepts such as, for example, causal explanation, induction, confirmation. Like other proceedings the volume provides information on current research projects or interests of the authors rather than on the subject. And it displays an equivalent printing quality and pricing: poor and expensive.
Humboldt
University,
Dagmar Berlin,
Grijppel Germany
Hans Jiirgen Seelos (ed.), W&terbuch der Medizinischen ~nfowrzatik (Walter de Gruyter, Berlin, 1990) XXII + 550 pp., 34 figures, hardcover, DM 148.00. The editor of this German ‘Dictionary of Medical Informatics’ seems to enertain a very broad concept of this discipline, one that not only encompasses medical information science but also medicine itself, nursing, law, technology, architecture, sociology, and industrial and health economics (p. VII). This strange understanding is mirrored in his choice of some 2000 terms of which you will find more or less useful definitions in this first edition of his multi-authored dictionary. Let us take a random look at the volume to see what kind of interesting entries we will find therein.