This research monograph is intended for researchers in AI and AI software development. It consists of seven chapters and two appendices; the first two chapters provide introductory material, the next four describe implementations of two AI problem-solving architectures, and the last chapter contains concluding remarks. The two appendices contain detail on an example refinement of an AI problem-solving specification.

Formalism is important, especially in AI architectures. The author is quite right in contributing this work, but a number of flaws will preclude its acceptance.

First, the title is somewhat misleading to computer scientists in general. The author is referring to problem solving architectures (paradigms). He concentrates on  CASSANDRA,  which is not widely known. In fact, the author admits to stopping work because of a lack of interest in the topic. He would do better to concentrate on blackboard architectures.

The text contains many typographical errors, starting with the first word of the preface. The hyphenation is awkward in many places. Neither the author nor the editor proofread the manuscript carefully.

In summary, this highly specialized text is not suited for a large audience, despite the claim on the cover. The general reader in computer science should not buy it.