![]() |
|||
![]()
|
![]() |
![]() |
![]() |
Chapter 3
|
1. | Introduction | |
1.1. | Representation of Knowledge: Putting the World into Computer | |
1.2. | Fundamental Tools: Verbal and Formal Mathematical Methods | |
2. | Background | |
2.1. | The Development of Logic | |
2.2. | Uncertainty | |
2.3. | Patterns | |
3. | Techniques, Practices, and Methodologies | |
3.1. | Conceptual Representation | |
3.2. | Relational Representation | |
3.3. | Representation of Uncertainty | |
3.4. | Open World Representation | |
3.5. | Connectivity, Patterns, and Schemes | |
3.6. | Representational Relations of Further Intelligent Actions | |
4. | Research Issues | |
5. | Future trends, Conclusions | |
References |
But I have learnt by trial of mankind
Mightier than deeds of puissance is the tongue
Sophocles: Philoctetes, 98-99, Transl.: F. Storr
From several points of view, representation of knowledge is the key problem of expert systems and of artificial intelligence (AI) in general. It is not by chance that one of the favorite namings of these products is knowledge-based systems. An expert, either a human or a machine, has her/his/its particular value by her/his/its knowledge. Though knowledge is always related to some special disciplines, professions, and/or experience, knowledge of any art has some generic features that define how this knowledge is acquisited, how it can be accessed, and how it can be applied to certain problems that are not totally identical to the earlier acquisited cases, i.e., learning, storing, retrieving, and reasoning methods based on knowledge. This circumscription covers somehow the concept of intelligence.
These generic features of knowledge are embodied in representation. A system, human or machine, in the process of learning (respectively input operations) stores the objects, actions, concepts, situations, and their relations in some representation form in the brain, respectively in the computer memory. This stored knowledge is used by retrieval (remembering), combination (association), and/or reasoning. If the computer was the same device as the human brain, the representation problem would be a biological and psychological task for achieving the best imitation. This is fortunately or unfortunately not the case; we can use only some analogies, metaphors between the brain and the computer; one should be always very careful as to how far these analogies work. On the other hand, the knowledge of the computer, which is really a directly or indirectly transferred human knowledge, is and should be some approximate copy of the human knowledge, transferred to another representation medium and by that way to another representation form.
Here we can estimate the relevance and difficulty of knowledge representation. Even the human origin is uncertain. The phenomena of the world, like the performance of a mechanical machine, of a living organism, the economy, or social relations are represented in the human brain by the signals of the sensory inputs like vision and touching, and by attaching those to representations of earlier inputs. Philosophy has hopelessly argued for nearly 3000 years over the relation of the world and of the sensory inputs, and other early developmental cognitive phenomena of the human and animal race. Where and how are all these stored, fitted together? Is it the brain only as a biological organism or something more, too, named mind or spirit, that joins this confusion? Here we are, after the sensory representation, only at the second phase of the representation process: the mental one. The third phase is the verbal representation. We experience clearly that the mental representation, e.g., an impression on a traffic situation, of a patient's status or anything else is not the same as the verbal representation. It is different: how we express it for ourselves, communicate with others, put down for a record. Even the nonwritten verbal representation and the written form are slightly different, even if we do not consider the representation means of metacommunication, like gestures, emphasis, pitch, etc.
This verbal representation should be transformed into computer programs, using the basic means of mathematical conceptual apparatus, which is much poorer than the rich human language. How this poverty is to be estimated is the main subject of modern linguistics, semantics, knowledge about the meaning of the words, semiotics about the general meaning problem of symbols, words included, and hermeneutics, the meaning of full texts in different individual, collective representations, like different social, generation, professional groups, or different cultures.
The representation problem is, by that way, a highly complex and highly practical one: by computer representation we try to surpass these dim relations, and get a practically applicable device. In application, we must have steady control over these relations, how the knowledge machine works really in interacting, influencing, governing the processes of reality. By this way the philosophical, biological, psychological, linguistic, and mathematical problems mentioned above receive a very practical relevance.
Our task is the presentation of the computer representation but it can never be detached from the previously exposed problems of representation phases; consequently, though the details of these previous phases are not the subject of this chapter, they will be referred to in all cases where it is unavoidable.
In this mirror, our representation task is a linguistic modeling work, the tool for intelligent actions, that should have:
Previous | Table of Contents | Next |
![]() |
|
Use of this site is subject certain Terms & Conditions. Copyright (c) 1996-1999 EarthWeb, Inc.. All rights reserved. Reproduction in whole or in part in any form or medium without express written permission of EarthWeb is prohibited. Please read our privacy policy for details. |