Knowledge Representation: Logical, Philosophical, and Computational Foundations

by John F. Sowa

John Sowa integrates logic, philosophy, linguistics, and computer science into this study of knowledge and its various models and implementations. His definitive new book shows how techniques of artificial intelligence, database design, and object-oriented programming help make knowledge explicit in a form that computer systems can use. The first three chapters are devoted to logic, ontology, and computable models of reality. Remaining chapters apply theories to the analysis of problems stated in ordinary language, and their translation to computable form. The text is self-contained, with each new idea defined when first mentioned; all formalism is developed in the body of the text or summarized in an appendix. Knowledge Representation is appropriate for advanced undergraduate and graduate students in computer science, as well as philosophy and linguistics students with some background in artificial intelligence or programming.

Genres: Philosophy, Computer Science, Logic, Mathematics, Nonfiction, Reference, Technical

512 pages, Paperback

First published August 17, 1999

Reviews

[Paul Black · Rating 4 out of 5]

A great text book. I had no idea that there was so much solid science/logic available. If you want to do knowledge right, you really should have a knowledge expert.

Some parts I like are unary, binary, and tirnary relations (page 61) and the 12 categories of "things": objects, processes, junctures, reasons, purposes, scripts, histories, etc. (page 75). There is a whole subsection of the limitations of logic: what in the real world can't logic describe? (page 357 et. seq.) Language patterns, or, questions and sentences to determine the characteristics of a word (page 445 et. seq.). For instance, the following make sense: a book, three books; an elephant, three elephants. The following phrases, also using nouns, sound odd or incorrect: a water, three waters; a happiness, three happinesses. The first are count nouns, the second are mass nouns.

[Hanna Abi Akl · Rating 5 out of 5]

Excellent introduction to logic, language and computation. John Sowa was ahead of his time and posed important concepts that continue to prove to be foundational in today's most pressing AI questions.

[Mikhail Filatov · Rating 3 out of 5]

A lot of interesting philosophical and historical details. But as an approach to AI it’s obviously outdated and the link between abstract concepts of predicate logic and relational databases, etc. is not very strongly established.

[Sanjay Sekar · Rating 5 out of 5]

One of the best books for people who are looking into implementing logical intuition in computer science.

[Groot · Rating 5 out of 5]

This is a wonderful book, deep and rich, historical and intellectual, written by someone who knows everything about the topics, and has spent a lifetime learning about, understanding and thinking about thinking. It's a graduate-level textbook, but is accessible to anyone from any field (assuming you score in the top 1% on any given standardized intelligence test, i.e., anyone reading this review).

[Jose · Rating 5 out of 5]

A must have for computer scientists. Explores both the information theory and its representation both from the computational and philosophical point of view.

It's not hard to read, and the history of the origins of Logic is very interesting and didactic.

The reader needn't to read the entire book, s/he can jump to the desire sections. Include exercises!

[Noah · Rating 5 out of 5]

Great overview of just what it says. This really puts ontology modeling, the semantic web, and reasoning technologies (e.g. AI) in context. Highly recommended but not for the non-geek or faint of heart. The prepositional logic appendix has a concise overview for review or picking up the syntax.

[Yasmina]

ok