The nature of the changes in information technology (IT) over the last decade shows that there is a shift from the data industry to the information and knowledge industry.
The results of the research show a cycle of 50 years after the change of the leading technology, in other words, the main innovation paradigm. According to the classification of analyst Norman Poiret of Merrill Lynch, there are six waves of paradigm shift. The previous waves were textile, railroad, automobile and computer. Now there is a “wave of distributed intelligence”. And the next wave is nanotechnology, biotechnology and possibly new types of energy.
It would be more correct to use the term “distributed intellect” rather than “distributed systems of information and knowledge provision“. Information is defined as a set of data and corresponding metadata, while knowledge is defined as information structured and provided in the most adequate form for the consumer (text, table, graph, sound or a set of them) depending on the context.
At present, IT has approached this critical point of “paradigm shift” from two sides – the information itself and from a technical point of view. According to Poiret, the paradigm of classical computing has already reached its maturity. But the development of distributed intelligence systems requires adequate computational and communication capabilities.
According to many experts, multicore and multithreaded solutions will gradually evolve into systems of distributed information. Sun Microsystems ideologist Greg Papodopoulos speaks about such systems in the context of Internet waves. Avalanche-like increase in the number of devices connected to the network, gradually transformed the classic network of computers in the network of specialized devices, and then the network of things. The basis of such networks will be cheap micro- and nanoprocessors capable of simultaneously processing many parallel data streams.
One more convincing argument of the possibility of implementing a new paradigm on the part of computational means (CT) is the development of GRID-systems. However, in IT sphere there are a number of problems which deter movement in the specified direction, namely:
- – orientation of information systems (IS) towards data and information rather than towards information and knowledge;
- – change of character of the software and hardware environment towards distribution is insufficiently supported by means of work with the distributed information convenient for end users;
- – unstructured public information and avalanche-like growth of its volume;
- – lack of generally accepted conceptual apparatus, which can be used by software tools for automatic information processing;
- – dependence of information on hardware-software environment leads to constant loss of information and knowledge at change of this environment, so it is impossible to store and increase them in electronic form for a long time;
- – location of a considerable part of the useful information;
- – change of character of problems solved and necessity not only wide use of means ICT, but also increase of their efficiency.
Thus, it is possible to ascertain, that IT are in a condition of change of information paradigm.
One of the forms of implementation of the new paradigm should become information spaces (IE) with the given properties.
Two opposite tendencies of the decision of arising problems are observed: scaling of the corporate centralized storehouses (or a database) and use of space the Internet for the decision of corporate problems. In other words, attempts to make a DBMS (or data warehouse) the size of the Internet from the Internet.
But in the first case the Internet space will lose the public access to the information with different service levels, in the second case the database will lose many local services.
The way out of this deadlock is to create IE of different scale with predefined properties. Such a corporate or a single IE, on the one hand, may have a very high service level, and on the other hand – a predefined compatibility with other IEs and the Internet space.
By IE we understand the information environment, which includes information sources and consumers, united within a corporation (which may be of any scale and be virtual or even conditional). FE must have any degree of nesting (up to a single computer) and scaling (without space limitations). Each individual FI may have some unique properties within itself, but must also be compatible with surrounding FIs.
The following basic information properties should be provided in a IP: distribution, notions, openness, heterogeneity, atomicity, interoperability.
To achieve the fundamental goals of IP it is necessary to solve the problems of multidimensionality of information and homogeneity of information entities.
The problems of multidimensionality are solved by separating each aspect into an independent information entity, e.g. content, structure, ontology, interconnections, assembly and use scenarios, etc. This information essence is formed as an independent information resource (IE) of this IP.
Due to the heterogeneity of the DI, in order to achieve the principal objectives it is necessary to make these resources homogeneous for this IP. For this purpose, the meta-tiers and meta-meta-tiers of the description of the DI in this PI are introduced sequentially. Thus, any information object can be described and atomized into multiple DI. In this PI, human being is also considered a resource.
The diversity of information objects in this IP is built by applying some resources to others. For example, the resource design is applied to the resource content and a publication is obtained, similar to a web page. In addition, we apply the resources of the class structure, get a publication of an electronic document.
Objects, classes, metaclasses and relations between them can act as a DI.
Traditional information tasks of design, management, etc. can be rethought in this DI. Many problems, for example through designing, simply cease to be such, and become ordinary problems of work with the information.
Concerning a basic element of IE, namely IE, the characteristic of a problem in terms of the target indicators defining level of its decision can be given as follows.
Distribution of the DI. The DI (and any aspect of it) can be located on more than one computer, presenting itself to the user as a single whole.
Notion of the DI. The presence of a connection of the DI with ontology elements for the purpose of automatic semantic processing of the resource.
Openness of the DI. Compliance with the principles of open IS. The degree of openness of the DI is determined by the availability of rules of formation and use of the DI for third-party users or applications.
Interoperability of the DI. Ability to use the DI from another hardware-software environment and transfer the DI to another environment.
Heterogeneity of the DI. Ability to use data of different nature and formats in a DI both when formalizing knowledge and when using them.
Multipurpose use of the DI. Possibility to use the DI, including knowledge, references to ontologies, etc., for different tasks.
Structural connectivity of different DI. Structural connectivity of different DI, including different expressive means (for example, structural connectivity of a flat drawing and text, three-dimensional model and table).
The basic role in construction of DI is played by systems of metadata which together with ontologies should provide semantic interpretation of information resources. As already existing metadata systems, the following can be mentioned:
- – Dublin Core – for general descriptions of PI;
- – CERIF-2004 – metadata for research information;
- – LOM (Learning Object Metadata) – the metadata of learning objects;
- – USMARC, RUSMARK – metadata for bibliographic descriptions;
- – HL7 (Health Level Seven) – Health Level Seven: a metadata standard for health information exchange and others.
The implementation of such a corporate IE will make it possible:
- – substantial cost savings in the creation of information and knowledge by reuse of previously obtained results;
- – a significant increase of the information and knowledge service due to a larger volume of available information, its relevance, speed of obtaining and processing;
- – new possibilities in information activity (designing, management, etc.) at the expense of possibility of processing of very large volumes of the information which are necessary for the decision of new on scale of problems;
- – continuity of results of information activity at the expense of long persistence and compatibility of these results;
- – occurrence of new technologies of information interaction between people, based on mobile devices and the distributed information;
- – the emergence of new ways of organizing various works and activities (collective, remote, distributed, virtual).
- – information integration of DI with other corporate IPs;
- – automatic integration of all IPs developed as part of such corporate IPs.
Advanced countries are now moving from the post-industrial stage to the stage of information society. Such transition is possible only through the widespread use of information and communication technologies (ICT) tools. But much less obvious is the evidence of a significant slowdown in efficiency gains from the use of ICT tools due to the fact that, with significant cost increases in this area, the usefulness (return) grows much slower.
This undesirable development can be reversed by developing and implementing the following directions:
- – ways of working with both structured data and raw data, unstructured and weakly structured;
- – metadata systems;
- – ontologies as a description of the system of concepts in the subject area;
- – ways and means of integrating different IPs;
- – agent-based approach as a way to create highly automated IPs;
- – theory and practice of corporate IPs;
- – service-oriented architecture;
- – semantic methods of information and knowledge processing;
- – ICT tools;
- – instrumental means in the above directions.
Essentially, it is a question of creation of a new infrastructure for paradigm of the distributed knowledge. It is a large and multidimensional work. It is impossible to see at once all problems and tasks connected with new technological breakthrough in the field of creation and use of distributed intelligence. But even the consideration and implementation of certain aspects of the new paradigm will certainly have a tangible effect. It is impossible to solve the task of implementing a new paradigm within a single theme or organization, but its implementation should bring an explosion of useful properties, functions and types of work with information.