24 von 27 Kunden fanden die folgende Rezension hilfreich
Dr. Lee D. Carlson
- Veröffentlicht auf Amazon.com
Format: Gebundene Ausgabe
The word `ontology' is usually associated with philosophical speculation on the reality of things, and if one checks the literature on philosophy one will find a diverse number of opinions on this reality. Engineers and scientists typically view philosophical musings on any topic as being impractical, and indulging oneself in these musings will cause one to lose sight of the topic or problem at hand. Rather than simplify the problem and make it understandable, philosophy tends in most cases to complicate it by endless debate on definitions and the use of sophisticated rhetoric that seems to have no bearing on the problem at hand. The conceptual spaces generated by these debates can become gigantic and therefore unwieldy, thus making the problem appear more complex than it actually is.
In the information age however, ontology has become a word that has taken on enormous practical significance. Business and scientific research are both areas that have increasingly relied on information technology not only to organize information but also to analyze data and make accurate predictions. In addition, financial constraints have forced many businesses to automate most of their internal processes, and this automation has brought about its own unique challenges. This push to automation usually involves being able to differentiate one thing from another, or one collection of data from another, or one concept from another. Thus one needs to think about questions of ontology, and this (very practical) need has brought about the rise of the field of `ontological engineering', which is the topic of this book.
The authors have given a good general overview of the different approaches to the creation of ontologies. There are many of them, some of which seem "natural", while others seem more esoteric. The reader though will obtain an objective discussion of the ontologies that the authors chose to include in the book. Discussions of the ones that are not included can readily be found on the Internet.
Given the plethora of ontologies that have been invented, it would be of interest to the ontological engineer to find common ground between them. The re-use of a particular ontology may be stymied by the different ontological commitments it is adhering to or it's actual content. In order to use it, it must therefore be "re-engineered". The authors discuss this prospect in the book, and define `ontological re-engineering' as the process where a conceptual model of an implemented ontology is transformed into one that is more suitable. The code in which the ontology is written is first reverse engineered, and then the conceptual model is reorganized into the new one. The new conceptual model is then implemented.
Also discussed in the book, and of enormous practical interest, is the automation of the ontology building process. Called `ontology learning' by the authors, they discuss a few of the ways in which this could take place. One of these methods concerns ontology learning using a `corpus of texts', and involves being able to distinguish between the `linguistic' and `conceptual' levels. Knowledge at the linguistic level is described in linguistic terms, while at the conceptual level in terms of concepts and the relations between them. Ontology learning is thus dependent on how the linguistic structures are exemplified in the conceptual level. Relations at the conceptual level for example could be extracted from sequences of words in the text that conform to a certain pattern. Another method comes from data mining and involves the use of association rules to find relations between concepts. The authors discuss two well-known methods for ontology learning from texts. Both of these methods are interesting in that they can apparently learn in contexts or environments that are not domain-specific. Being able to learn over different domains is very important from the standpoint of the artificial intelligence community and these methods are a step in that direction. The processes of `alignment', `merging', and `cooperative construction' of ontologies that are discussed in the book are also of great interest in artificial intelligence, since they too will be of assistance in the attempt to design a machine that can reason over multiple domains.
The ontologies that are actually built are of course not unique. This results in a kind of semantic or cognitive relativism between the environments that might be built on different ontologies, even in the same domain. Merging and alignment both address this relativism, along with other techniques that are discussed in the book. The selection of the actual language that is used to create an ontology is also somewhat arbitrary. The authors devote a fair amount of space in the book to the different languages that have been used to build ontologies. Through an elementary example, they discuss eleven different languages, namely KIF, Ontolingua, LOOM, OCML, Flogic, SHOE, XOL, RDF(S), OIL, DAML+OIL, and OWL. The choice of a language is dictated by what one is seeking in terms of `expressiveness' and what kind of reasoning patterns are to be deployed when using the ontology. The authors point to a tradeoff between the expressive power of the language and the reasoning patterns that are attached to the language. The expressiveness of a language is directly proportional to the complexity of the reasoning patterns that are used.
Ontological engineering as it presently exists is still carried out by a human engineer. To create an ontology every time from scratch would be tedious, and so it is no surprise that tools were invented to make ontology creation more straightforward. Some of these tools are discussed in the book, such as KAON, OilEd, Ontolingua, OntoSaurus, Protege-2000, WebODE, and WebOnto, along with assessments as to their utility. The discussion is helpful for newcomers to ontological engineering who need guidance as to what direction to take. The automation of ontology building would of course be a major advance. To accomplish this however would require that the machine be able to simultaneously and recursively construct the knowledge base and reason over it effectively. This is a formidable challenge indeed.