"New" Organization of Knowledge and the Information Science

Jela Steinerová

Introduction

The Internet and information technologies are changing the starting positions of the library and information science. From the reception of information point of view, digitalized information and the media on the one hand provide for a better access to information, but, on the other, they increase the chaos of disorganized information. For that reason, the importance of a scientific discipline that searches for the optimum methods of structuring recorded information has been increasing. In fact, it tries to organize knowledge in the broadest meaning of the word. The development of organization of knowledge indicates a gradual shift from traditional hierarchic structures (e.g. grading systems) to a more loosely structured indexed systems, dictionaries and thesauruses, up to full texts, delinearized text structures (e.g. hypertexts), and non-traditional models and meta-languages, a general representational tool derived from artificial intelligence (e.g. associative networks, production rules, procedural methods, frames, etc.).

In this article, we will attempt to highlight the significance of the method of organizing knowledge for the purposes of information science and for the practice of information-oriented institutions. We will identify the common links between reality, understanding, and their recording, expression and use in specific situations involving information transfer in society. The specific starting point will be the semiotic properties of knowledge. The basic question is one of interpretation and representation of knowledge and objects. Do the established laws of the interpretation theory still hold true in the new "digitalized" space (or "cyberspace" - to use M. Caskells' terminology)? What are the characteristics of the basic groups of users in this space? Do their expectations, needs and requirements vary? What is, in fact, the man's position in the "new" organization of knowledge?

1 Physical and Intellectual Approach to Information and Information Sources

The basic semiotic change of the traditional library concept arises from the juxtaposition of objects existing in reality and their images (representation). The physical access to holdings, which used to be determined by the location of documents, is being abandoned in favour of an intellectual access which focuses on the utilization of information in digital form (e.g. digital library). How does it affect the traditional ways of organizing knowledge? It differs particularly in the emphasis on the support of selection, analysis, synthesis, interpretation and the presentation from the sources. A simple navigation or identification of source is no longer good enough, since for today's users, secondary information without interpretation or analysis are an insufficient output of information systems and services. Thus we speak about a "new" organization of knowledge in accordance with the trends of a "new" artificial intelligence.

We define organization of knowledge as a set of methods for imposing a structure on reality with the intent to capture the ontological content of the structures of knowledge. Since the time of Ranganathan we have been trying to structure reality by means of various categories of existence. For example, Ranganathan's facet classification is based on the following ontological categories: time, space, matter, energy, personality and subject area. These categories have been elaborated upon and used in a number of tools and models of cognition up to the present time. In particular, representation of knowledge means creating an image of reality (object) while capturing its function, and the semantic relationship between it and its image. It is therefore the case of an agreed-upon abstract type structure of cognition and, concomitantly, a process of a description of the structure's development.

The intellectual approach to information sources emphasizes information as a psychic and social construct linked to the man's personality and its development (genetic or historic). Information is not a finished entity or product, but rather a process of creating meanings. It is a typically human process and reflects the cognitive, emotional, social, and physical components of human personality.

If we accept such a non-traditional concept of information services in libraries or other institutions, we also have to search for more effective tools to organize the body of knowledge that would more naturally support the creative process of utilizing information, in contrast to the existing and nowadays used information languages. An entire range of models and languages for representation of learning in intelligent systems falls into this category. The current models and meta-models are capable of evolution, they represent certain dynamics of processes of information use, and they register context, change and individual situation.

2 Characteristics of the New Knowledge-Organizing Tools

What characteristics are then required from new tools for the organization of knowledge today? First of all, the ability to articulate concepts and their meanings, the ability to determine their fitness for larger categories (especially mutual relations) and the ability to express their utility in specific situations. They are ideally combined in semantic, syntactic and pragmatic aspects of semiotics. An example of such a solution may be the emergence of a frame representation of cognition, as conceived by Marvin Minsky.

However, even those aspects did not suffice to express reality and to provide the best representation of man's perceptions and his thought process. The new artificial intelligence introduces the ideas of narrowing the gap between the natural (human) imagery (representations) and artificial systems of knowledge representation. As of now, their meeting or fusion cannot be reliably predicted. Will it be the case of virtual reality penetrating into real life - or the man's adaptation to technology by gradual integration of the natural (human) and the artificial? The "new" artificial intelligence is in fact shifting from the investigation of individual cognitive processes to the social and cultural aspects of the knowledge functioning as to a context which, in effect, determines how the objects of reality behave within systems.

The requirements for an effective and accurate organization and representation of knowledge have to do with the concept of intelligent retrieval. Such retrieval may be described as an information retrieval utilizing the results from the cognitive science and artificial intelligence. The "new" organization of knowledge and the "new" information science are in this case "nourished" from the "fountain" of a changed paradigm of theoretical thinking about information processing models and their simulation in artificial systems.

As these disciplines evolve, the semantic organization of knowledge in networks takes on greater and greater importance. The meaning of words and the meaning of relationships between them form a framework for a contextual application of information. In the course of search and the organization of knowledge, an active interaction of knowledge structures by means of change control and feedback becomes increasingly feasible. Consequently, the user may review multiple meanings and broaden the search accordingly. That is why the new organization of knowledge focuses primarily on accuracy, multimedia capability, transparency and simplicity of terms in formulating queries. In fact it moves closer to the natural communicative and cognitive patterns. Given more and more open access to information, the manner in which knowledge is organized should help select and eliminate inappropriate data. Interconnectivity of the various forms of information is related to the uniform approach to the search for it. A multimedia capability and a simple approach to organized knowledge are thus emerging even in the area of organization of knowledge.

Technology gradually overcomes thwue physical limitations of accessing information sources. The intellectual approach is based on better understanding of cognitive aspects of learning and their expression in the new conceptual structures. For example, cognitive science investigates those aspects that lead to the formation of concepts, such as generators and inhibitors of thinking that depend on personality types. Associations (the most visible and empirically verifiable) were the first ones identified. They form the foundations of the associative methods in the languages of knowledge representation, particularly the associative networks, which may be divided into semantic networks, causal networks and neuron networks, and possibly also connection models. This type of knowledge organization also gave rise to the so-called associative (conceptual) retrieval.

Aside from associations, however, there are other possible generators and inhibitors of thinking which affect the treatment of concepts. Among those are especially analogies, metaphors, practical activity models, first information, domination and other, hitherto undiscovered aspects of natural human thinking. Their simulation by means of artificial intelligence systems is helpful in solving multiple questions of organization of knowledge and of information retrieval. As far as the cognitive science is concerned, of the greatest utility in organizing knowledge are the processes of categorization and concept formalization. To some extent though, these processes are subjective in that they depend not only on the personality type and the education level, but also on the types of concept categorization deployed in thinking. At least two such types have been differentiated thus far: one with a predominance of practical manipulation of objects, and the other with the preponderance of concept categorization by common traits of certain objects.

The intellectual approach to information opens new problems in the organization of knowledge, particularly where electronic information is concerned. The design of knowledge organization systems has to recognize the principles of uncertainty, variety and complexity, which characterize the human learning processes. The most important is the determination of meaningful fragments of knowledge and their mutual relations. In terms of identification, this is a question of making decisions about complex documents that are comprised of several parts. The key issue here is the identification of the parts, and the establishment of mutual relationships of parts to each other and to the whole. Another difficult question is the identification of changes (modifications) in the individual parts. In this context, the possibilities of conceptual representation of names (often in a certain hierarchy) and of the measure of contextual information in these names (titles) are being explored.

3 Modelling and Organization of Knowledge

The traditional languages of knowledge organization, as applied to books, model reality with the objective of attaching to documents some information about those documents. The present-day cataloguing and classification procedures are finding new applications also in the area of electronic information processing (e.g. in document management systems, in the Internet - web applications and in digital libraries). A more modern terminology refers to these data as meta-information about digitized objects. It enables description, organization, "discovery" and access to information systems in the networks. This brings them increasingly closer to methods of modelling knowledge and objects of reality. For that reason we need to take a closer look at the possibilities of modelling knowledge in the connection of searching for newer methods of knowledge organization.

Model can be defined as a representation (projection) of a segment of reality conditioned by the human cognitive activity. It the narrower (application / technological) sense, the model is a representation system that establishes relations, an interface between the man and the system. Every model has to acknowledge that the observations have to be expressed (via meanings of concepts), engaged into relationships with other observations, and be useable, earmarked for specific purposes. The semiotics (general linguistic, the science of symbols) has for some time recognized the following three aspects of symbols: semantics, syntax and pragmatics. In modelling of complex social problems including knowledge or objects of reality, these aspects are evident also in the individual pieces of meta-information.

From the standpoint of semantics, we are dealing with individual properties (attributes) of objects and observations in the case of meta-information as a model. In library systems, the sets of those properties are defined in advance in the well-known standards and formats (such as the MARC and its specific applications). As opposed to the Internet (web applications) and the document management in the case of teamwork, the scope of semantic data (embodied in the description of the document properties) is the most extensively developed in the universality of tools for cataloguing and classification. In the process of detailing, as far as modelling is concerned, these tools approach the methods of object-oriented modelling, particularly in terms of sorting into certain classes of objects (e.g. documents) with their properties, heritage functions, specialization, or example generation.

The syntax aspect is, in essence, connected with the structure and structuring of knowledge by means of concepts. It is necessary to determine the type of relationship between the individual elements of knowledge in advance (in the case of object-oriented methods they are frequently conditioned by the ontological structure of knowledge, which models reality through its functional components and their mutual relationships). The issues of relationships between object properties and elements of knowledge have not been sufficiently developed in the traditional library tools for organizing knowledge (e.g. MARC). Here, the new organization of knowledge leaves something to be desired.

The pragmatic aspects of knowledge modelling are much the same. The problem is how to express the link between concepts and concrete situations of their use, with concrete groups of users. Some cognitive methods focus on the multidimensional nature of knowledge. They strive to foresee and represent, ahead of time, various potential meanings, various potential situations and interpretations. This principle has resulted in so-called polyrepresentation of knowledge. The question remains, though, if we can define, a priori, all situations that trigger a concrete representation and interpretation. Here the new organization of knowledge is again confronted by the problem of better understanding of human thinking and the cognitive activity. However, the cognitive science models have always indicated that the logical and intuitive activity of human thinking is inseparably intertwined with the emotional and non-conscious components, and that the contribution of the emotional element to the formation and utilization of knowledge is probably greater than originally thought. That complicates the attempts to come up with realistic learning models. On the other hand, it constitutes a challenge for both the cognitive and the information sciences.

4 Digital Libraries and Organization of Knowledge

Aside from the basic problems of knowledge modelling, the digital libraries and broader manipulation with digital objects brings about some practical problems dealing with a unique identification of the author or the person connected with the life cycle of a digital object. Another acute problem surfaces when we try to define the "permitted" operations on an individual objects (in case of digital libraries research, mainly the copyright and the transactions associated with the use of and access to digital objects). Again, it is apparent that, in real life, technological possibilities are dominated by the social (and organizational) rules of the utilization of digital objects. From the point of view of functionalities, rules are being overtaken by technology, which are developing much faster. Not all methods (from the point of view of, e.g., copyright) are unequivocal and easily transferable from their "physical" environment (of, e.g. traditional libraries) into the electronic environment. Therefore, more attention should be given to the methods of organizing knowledge in that particular environment. The information science has the task of combining the cognitive and social (communicative) procedures, especially in regards to applications of information technologies.

The problem of knowledge organization in digital libraries, document management (knowledge management) and web documents is the representation of the intellectual content (of the information). Digital libraries, for example, contain an enormous number of forms, media and formats that should not impact the organization of knowledge itself. The essential is the content (knowledge), its preservation and the prospect of its use. While the problem of long-term preservation may be related to the "survival" capabilities of the medium used, the length of, and good prospects for, the use of instruments for knowledge organization are equally as important. In this sense, the web systems are lacking the organizational quality with added value (analysis, synthesis, interpretation). That is the reason why second generation WWW systems are being developed among other things in research laboratories at American universities, which emphasize the logical and semantic aspects of a quantity of heterogeneous data. These are mainly embodied in the representation of context (or contexts), individualized and group use models, as well as ontological representations of the environment.

Research of digital libraries also shows that social and human factors are going to play a more and more important role in them. The search function is gradually shifted to a juxtaposition of the function of the true "finding" of relevant information. The technology of cognitive (intelligent) agents as a realistic application of artificial intelligence and the cognitive science is being applied particularly in natural human processes of support in the resolution of difficult problems ("advisors" or "consultants" in the decision-taking in various special situations and applications, in resolution of problems or in the area of information retrieval). That means that even the new web architecture, in contradiction to the well-known principle of physical hypertext link, is moving in the direction of information management, manipulation and use on the logical and the semantic levels.

5 Possibilities of Cognitive Models

Cognitive science presents stimuli through the well-developed models of perception, memory and thinking. For example the "Whole Brain/Four Quadrant Model" (Ned Herrmann) draws on the traditional division of the brain functions between the right and left hemisphere, with the cognitive and intellectual aspects, including personal preferences, and structural and emotional elements. In addition, a "domination" principle of certain functions is recognized. This model interprets, metaphorically, the approach to thinking and preferential ways of cognition. The first approach, A, is the upper left (cerebral) quadrant, characterized by analytical, mathematical, and technical way of thinking and problem solving. The second approach, B, is the lower left (limbic) component, associated with control, conservativeness, planning, organization and managing (administration). The third approach, C, refers to the lower right (limbic) component, characterized by interpersonal, emotional and spiritual modes, and the "narrator's" functions. The upper right cerebral component (D) prefers imagination, synthesis, as well as artistic, holistic and conceptual ways of thinking. This example clearly shows that the cognitive science approaches the issues of knowledge models in a holistic manner, and that it provides inspiration for the information science, most notably in the area of knowledge systemization.

Contemporary models of knowledge systemization, particularly practical models of knowledge, model individual cognitive agents (solving problems). The majority of traditional models ignore the social and organizational contexts, not to mention emotionality. Consequently, some representational frameworks for alternative, even conflicting, viewpoints in certain defined domains are being sought. In modelling, the description (representation) of these viewpoints is organized into a hierarchy that enables the users to understand the views of other users and to articulate their own cognitive/social context.

The basic arguments casting doubt on the possibility of simulating the natural human thought processes in the perception models stem from the principles of conscience (self-realization, subjectivity), autonomy, intentionality (directionality, goal-oriented nature of human thinking and learning) and unity (the whole is more than a sum of individual parts). All these qualities are interconnected by a meta-plane of human cognitive activity, of which not much is known. The meta-plane was mentioned for the first time in works discussing the relationship of language and thinking (Wittgenstein). Similarly, it is difficult to model the value-based aspect of learning (by virtue of its organization), creativity, imagination, motivation and emotionality. However, when the new models of knowledge succeed in transcending the original deterministic concepts (as is now being successfully realized), we will be able to contemplate a new organization of knowledge that will be adjustable to the natural characteristics of human perception. It will be dynamic, capable of organizing incompleteness, inaccuracy, ambiguity and open-endedness. It will be capable of evolving with the social, cultural, and organizational character of learning, supporting interactivity and cooperativeness as well as the value-based relation of knowledge to reality.

6 Changes of Interpretation and Representation in Digitalized Space.

It appears that the technological digital revolution permits to create some new "artificial" worlds. These worlds probably offer something else in addition to a simple simulation or reflection of our existing world. For example, today we do not have a definitive answer to the question of what good would it be to have an (artificial) feeling-recognition equipment, although work on such systems is under way and they slowly start to come out of research laboratories. It is therefore reasonable to assume that this development will bring about some changes in the possibilities of interpretation and representation of digitalized inputs as well.

The interpretation (of knowledge) is a process of assigning meaning to entities (objects, events, activities), in accordance with the specific context, mainly the situation in which the subject happens to be. Even in the digital library environment alone, it seems obvious that the possibilities of interpretation and representation are much richer than in a traditional library. The growth is fairly evolutionary, meaning that the newer tools for access and organization of knowledge do not substitute the traditional book sources but rather integrate them into a new quality. The digital library is almost a "hybrid library" in the sense that it integrates the administration and organization of the physical place with the administration and organization of a digital space.

The access to information objects and to their interpretation is speedy, dynamic (interpretation changes of information in time are accepted) and it integrates directly with the sources. The development of knowledge organization could thus be seen in the gradual integration of interpretation into its tools, as well as into the sources of information. Organization of knowledge must be adapted to the user. Its system should enable the user model to connect with the model of available sources via the new interpretative access mechanisms (agents). The interpretation-supporting system could approximate the cognitive creativity-supporting agents by evaluating a number of existing creative solutions to problems and deriving from them the suggestions for solving the new ones.

The digital space opens new possibilities of knowledge organization while capturing multiple representations of reality. In practice, this appears as a "multi-thesaurus manager" or as a guide to structured terminology. The most significant is the fact that the knowledge so organized accepts development, change, and makes even preliminary "self-learning" of knowledge organization structures possible. Once the problems of access, speed or immediacy of contact with the approximation of the natural character of human communication (sic) are solved, then more space can be devoted to "intellectually" more demanding aspects of knowledge organization, namely to the interpretation processes and their support.

The organization of knowledge in digital libraries approaches the interpretation aspects by means of questions designed to scrutinize people, partly as individual users, and partly as members of groups and communities, producers, users and managers of information who keep learning. The key problem, for purposes of organization of knowledge, is the recognition of priority functions of knowledge for certain user communities, as well as the recognition of individual cognitive processes that support data interpretation and utilization. The organization of knowledge in a digital space, just as generation, utilization and interpretation of information, is socially conditioned activity. It appears that the successes in creating user interfaces will be conducive to the development of user-flexible knowledge organization systems. Consideration must be given to the overlapping of human functions in manipulating and organizing knowledge, from producers to consumers and their respective groupings. For example, a scientific researcher will require a different interpretation than a small child, although the knowledge domain will remain the same. The same applies to legislative documents (legal experts' needs vs. layman's need).

7 New Functions of Organization of Knowledge

Another property of knowledge organization in this sense is learning to work with its methods and the support of these learning functions by the systems (in digital libraries). Authors of information will probably need other, newer supporting tools featuring integral (general) functions of knowledge organization when communicating meanings in groups. We also have to recognize the different levels of generality in representation and interpretation of knowledge, the requirements of dynamics, transferability, the richness of relations between the various objects of information. The organization of knowledge should also "take over" the functions of filtering and selecting information, of navigation, as well as that of the primary analysis and presentation. New forms of digital network communication (such as chat rooms and supporting consulting systems or networks) bring about a new "culture" in knowledge systemization. Within this space, the products and systems that are "custom-made" begin to take on importance. That's why the knowledge organization should "behave" accordingly and accept the differences in needs, as for example between academia and common users.

The knowledge organization in "newer" applications tends to converge almost dangerously with the tools of knowledge representation in artificial intelligence.

The dominant factor is the human thinking and cognition. On the basis of this understanding, virtual environments with the supportive functions in solving problems are being designed. Cognitive and behavioural experience of the users is incorporated into the instruments of information retrieval and utilization support.

An example of the cognitive styles is the division into the theme-dependent and the theme-independent, or an explicit division into inexperienced (beginning) users and experienced users (in terms of technology manipulation or the skills in using the information search).

Another classification, from the standpoint of learning styles and subsequent behavioural patterns, makes a distinction between holists (who approach problems from the point of view of the entire whole) and serialists (approaching problems by interconnecting individual specific parts). On this basis we can deduce that one group of users prefers concept maps in knowledge organization that reflect the whole with its relationships (holists), while the other group has a preference for lists of key words from which they construct their own representation of the problem (serialists). Therefore, the organization of knowledge has to be flexible at least in self-organization of some fundamental cognitive and behavioural functions. Classification methods based on artificial neuron networks keep drawing closer and closer to this ideal.

A principle change in the interpretation and representation, in contrast to the traditional librarian approach, lies in the fact that the relationship between reality and interpretation is more resilient, it supports variability of interpretation in the individual phases and the changes of cognitive states in the course of solving problems. The organization of knowledge enables to support complex intellectual processes tied to the behaviour of dissimilar groups searching for information classified according to its cognitive or social characteristics.

Consequently, the organization of knowledge should support even those portions of natural human thinking and learning that have been considered "taboo" and which do not have to be burdened by the tradition of library and bibliographic systemization. The predominant scientific methodology, the logic and rationalism in the philosophical sense of the word, have served to limit the process of modelling and utilizing knowledge. They have suppressed the natural human and social elements, intentionally simplified and reduced the problems to superficial observable phenomena. Today, the situation is quite different. Thanks also to progress in technology, we are now capable of supporting and modelling even natural human and social aspects of man's contact with information. An appropriate example may be the support of human creativity (not only when searching and using information). The traditional vision of creativity as a divine "inspiration", or an exclusively human activity impossible to model, is being gradually overcome. Creativity is looked upon as the ability to metamorphose known stimuli and make them in new combinations, or, conversely, the ability to examine and transform conceptual spaces. In both cases, the modelling may take advantage of the modern technological tools (cognitive science research offers examples from the fields of music, architecture and graphic art).

Conclusion

The "new" organization of knowledge proves that the library and information sciences come increasingly closer to psychology, philosophy and sociology. Of these disciplines they, however, select those knowledge, methods and instruments that explain the complex relationships between man, information and information technologies, as well as the processes of the immediate as well as mediated communication of information in society. In that the information science resembles the newer interdisciplinary fields along the border between technical and social sciences, such as, e.g., the cognitive science and artificial intelligence. This scientific discipline cannot be reduced to its practical institutional component (libraries in the traditional sense of the word). In that regard, the concept and the practical execution of the digital library proves that it is more and more important to understand the richness of social relations and cognitive processes contained in the generation and utilization of data in any form, in any space or time, dynamics and flexibility.

It appears that the current organization of knowledge cannot keep up with the rapid technological advances. The traditional library methods need to be enriched and developed by means of new modern methods of knowledge modelling. Their common qualities manifest themselves in depiction of the object, its properties, in multiple dimensions and in the relationship of an object to other objects and their representations. That is outlined in Fig. 1.

REPRESENTATION	OBJECT	PROPERTIES
MODEL	RELATIONS	STATES
		SITUATIONS DEVELOPMENT

Fig. 1. Common properties of the newer tools for the organization and modelling of knowledge

Further, it should be kept in mind that the organization of knowledge should be directed to functionally useful support of human decision-making, not only to static abstract models. In this sense the organization of knowledge approaches the concepts of artificial intelligence in constructing agents that support creative problem-solving in concrete situations. They are in effect systems (knowledge models) that serve as intermediaries with special skills. That is why it is necessary to make a shift from universalism in knowledge organization to those situations that may be useful and necessary in social interaction and communication of knowledge. The representation of knowledge, aside from interpretation, thus connects more and more with activity, with doing, with the cultural conditioning of symbols representing knowledge. In the area of artificial intelligence, the methods of decentralization and behavioural programming (communities of reactive components) are being applied. There is a gradual transition from properties of knowledge on an individual level to a collective level, tied to the environment and anchored in reality. The knowledge organization should gradually add these and other recognized factors of human knowledge to its toolbox, including analogy and metaphor.

In the future, several scientific and practical disciplines will influence the development of knowledge organization. In addition to artificial intelligence and cognitive science, there is the philosophy of the mind, the theory of interpretation, semiotics, social communication and interactionism. In any case, the knowledge organization, and the information science with it, is emerging from the "dusty" schemes and turning in the direction of man and his relationships in practical problem situations. The knowledge systemization could find inspiration in the rediscovered theories of personality (Freud, Jung, Fromm) and in the results of cognitive psychology.

For example, it is necessary to come to terms with the notion of equal influence of both cerebral hemispheres in cognitive functions. Up to now, the preference in education and in the organization of knowledge was given to verbal and analytical thinking (left hemisphere) as opposed to the non-verbal and global thinking (the right side). Newer research points out the speed, complexity, and integrity of the spatial and perceptional processing in the right side of the brain, and the manner in which both hemispheres complement each other in handling information within the human mind.

For that reason it might be inspirational to consider the concept of duality even in the organization of knowledge, drawing on the natural division of the parallel ways of cognition into the rational and intuitive, convergence and divergence, abstract and concrete, controlled and free, linear and non-linear, analytic and holistic, objective and subjective, sequential and simultaneous, all the way to the duality of Ying and Yang (male and female) and the consequential opposites such as negative and positive, darkness and light, heat and cold, subconscious and conscious, submissive and aggressive, etc.

At the present time, the possibilities of using a variety of stimuli to inspire the organization of knowledge are getting greater and greater. The philosophical foundations overlap with technological capabilities and for those of us working in the information field, the challenge remains to contribute to this research with the methods, observations and experience from the library tradition. By collaborating with other information-oriented professions, we can help create a new, modern and effective organization of knowledge.

Literature

[1] Ricoeur, Paul. Teória interpretácie : diskurz a prebytok významu. Bratislava : Archa, 1997. 135 s. ISBN 80-7115-101-7.
[2] Steinerová, Jela. Teória informačného prieskumu. Bratislava: SlTK-CVTI, 1996. 262 s. ISBN 80-85165-58-9.
[3] Minsky, Marvin. Konštrukcia mysle. Bratislava : Archa, 1996. 103 s. ISBN 80-7115-107-6
[4] Kelemen, Jozef. Strojovia a agenty. Bratislava : Archa, 1994.109 s. ISBN 80-7115-089-4.
[5] Steinerová, Jela. Tvorba informačných produktov : Nové prístupy informačnej vedy. Bratislava : CVTI, 1998. 130 s. ISBN 80-85165-73-2
[6] Steinerová, Jela. Význam informačnej analýzy v informačnej spoločnosti. In: Zborník FiFUK. Knižničná a informačná veda. 17. Bratislava : UK, 1998, s. 37-58.
[7] Steinerová, Jela. Interpreting Knowledge in Information Science. In: Knowledge Management und Kommunikationssysteme. ISI´98. Hrsg. H. H. Zimmermann, V. Schramm. Konstanz : UVK, Univ.-Verl. Konstanz, 1998, s. 143-153.
[8] Pstružina, K. Svět poznávání : K filozofickým základům kognitivní vědy. Olomouc : Nakladatelství Olomouc, 1998. 183 s. ISBN 80-7182-074-1.
[9] Borgman, C. et al. UCLA-NSF Social Aspects of Digital Libraries Workshop: Final Report. http://www-lis.gseis.ucla.edu/DL/UCLA_DL_Report.html