CONTENTS: Reviews, written by a staff of over 12,000 expert reviewers, of current mathematical literature (books and journal articles) of research interest.

FREQUENCY Monthly (supplemented by annual indexes).

COVERAGE: The world literature of mathematical research. At present, about 54,000 items (books, collections and journal articles) are reviewed each year.

FUNCTION: An indispensable tool for scholars and

research workers in mathematics and the mathematical sciences.

CIRCULATION: Approximately 850 paid. The list price for a 2001 subscription to MR, including the annual index, is $6,587 = $6,141 (Data Access Fee) + $446 (Product Delivery Fee for MR paper). Most of these are institutional subscriptions (libraries, research institutes, government and industrial laboratories) and therefore have multiple readership. The MR mailing list is maintained at the headquarters of the American Mathematical Society in Providence, Rhode Island, USA.  

MATHSCI, the electronic version of CMP and MR, contains all the bibliographic data given in CMP and MR since 1940 and the entire text of the reviews in MR since 1940. It is available in several formats, each of which can be used to find items by searching on, among others, author names, titles, review texts and classifications. MathSciNet (www.ams.org/mathscinet) is the World Wide Web version of MathSci. MathSci Online, which contains additional listings in statistics and computer science, is available through several commercial vendors. The MathSci data is also available in CD-ROM format in a multi-disc set called MathSci Disc.  

READERSHIP: Through subscriptions and exchange agreements, MR is available to the international community of mathematicians and other mathematical scientists. MR is published by the American Mathematical Society. PO Box 6248, Providence, RI 02940-6248. USA.

Address all editorial correspondence to

 

Mathematical Reviews

416 Fourth Street

PO Box 8604

Ann Arbor, MI 48107-8604

USA

Editorial correspondence can also be sent electronically to

MATHREV@AMS.ORG

or bv FAX to (734) 996-2916.

2001d:68213 68T37 03B48 68-02

Pearl, Judea (1-UCLA-CG; Los Angeles, CA)

*Causality. (English summary)

Models, reasoning, and inference.

Cambridge University Press, Cambridge, 2000. xvi+384 pp.

$39.95. ISBN 0-521-77362-8

In this book the author provides a comprehensive and incredibly

lucid account of probabilistic and graphical models for causality

and describes how the representation of causal information can allow feasible reasoning about complex situations.

The introductory section describes how causal networks can be modelled using the notion of acyclic graphs. This approach is then

related to functional causal models consisting of structural equations. Also discussed is the effect of actions on causal networks and how networks can be updated in the light of new information. Later sections describe algorithms for learning networks from data based, for example, on notions of model preference and Occam's razor.

Overall this book is a rich contribution to the areas of statistics and artificial intelligence. It describes ideas at the cutting edge of research into causal networks while being extraordinarily accessible.

               Jonathan Lawry (4-BRST-EM; Bristol)