Many in life have had to deal with people who seek only to obtain benefits. Moral and other aspects of life are of secondary importance to them.

Views, beliefs and actions are directed solely to obtaining results that are useful in a practical sense. Those around him often condemn him for this.

Immediacy and artlessness in the eyes of a pragmatist is stupidity.
Ilya Nikolaevich Shevelev

Pragmatic thinking style

Pragmatists strive to achieve the goal, using all the possibilities now available. They will not look for additional information, funds, resources, because this is an unjustified loss of time and effort. Problems are solved as they arise, so as not to be distracted by the main goal - obtaining a specific result, even a small one.

The constant search for new methods, experiments and other activities do not indicate a deviation from the chosen course. This does not come from the desire for novelty, but is dictated by the desire to quickly achieve results. For the sake of this, they are ready to listen to someone else's opinion, in the hope of finding the shortest path to the goal.

Such an approach may seem superficial. It differs from generally accepted norms, and pragmatists give the impression of inconsistent, principleless people. They are of the opinion that everything that happens around depends little on the abilities and desires of a person. The main thing for pragmatists is not to miss the favorable moment when everything is going well. Their belief in the unpredictability and uncontrollability of the world justifies the strategy "today it will be like this, and then according to circumstances."

It is impossible to influence a pragmatist with emotions and manifestations of feelings, unless they become an objective obstacle on the way or, on the contrary, help in a given situation. They perfectly feel the conjuncture, quickly responding to its changes. They easily cooperate, enthusiastically participating in discussions of important issues and the development of collective decisions.

Pessimism, a negative attitude is not characteristic of these people. The problems that arise are not able to turn them off the chosen path. They are connected to the decision with a positive attitude, a pragmatist, in simple words, an incorrigible optimist who seeks to turn difficult circumstances in his favor. The folded worldview does not allow to overdramatize and take the difficulties that arise too seriously.

Behavior and thinking are flexible. Communication skills are well developed, they can easily imagine themselves in the place of another person and understand the consequences of their actions. They are not indifferent to someone else's opinion exactly to the extent that their future depends on it.

Features of the behavior of a pragmatist

Pragmatic people often achieve success in politics and management. This is due to their character, life attitudes, style of thinking.

They are characterized by:

• search for the shortest paths to profit;
• quick adaptation to new conditions;
• interest in new methods, innovations;
• use of any means to achieve goals;
• creativity.

They are intelligent, learn new things quickly, using every opportunity to get closer to the intended goal.

Management values ​​pragmatists for the following qualities:

• concentration on obtaining maximum profit, the fastest return on investment;
• pre-thinking tactical and strategic aspects of the case;
• the ability to influence others, to convince them of the correctness of their ideas;
• not lost in difficult situations, looking for non-standard ways out of them;
• loves bold experiments, introduces innovations.

Cons of pragmatism

Like all other people, pragmatists have not only strengths, but also weaknesses.

They appear as:

• indifference to the distant prospects of the business, which in the near future will not bring income;
• the desire to achieve early results at any cost, a long wait is not in their nature;
• attention is focused only on the material side of the matter, everything else does not matter;
• from the outside it seems that for the sake of profit they are ready for any compromises;
• tendencies to maximalism, of all available resources, they try to get the greatest return.

Pragmatists won't worry about failure for long. They will look for new ways if the old methods no longer work. Having drawn conclusions for themselves from the mistakes made, they will not repeat them in the future.
They understand that it takes a lot of work to reach their goal.

They will not rely on outside support, they are used to relying only on themselves. They can help if you ask them to. If in the future there is an opportunity to compensate for the costs, then the applicant's chances increase significantly.

Inactivity is impossible for them, a pragmatist is a person who, with his optimism, is able to inspire others to labor achievements. Developed intuition allows you to choose from a variety of options one, but effective and quickly giving returns.

A cynic, a romantic, a lyricist, a pragmatist - absolutely everyone dreams that someday "scarlet sails" will appear on their life horizon.
Oleg Roy

Pragmatist and relationships with others

In communication with others, a pragmatic person makes a good impression. He is open to communication, likes to joke, does not argue, easily finds contact with any people. In conversation, he often uses examples from life, stereotyped phrases. The tone of statements is often enthusiastic, enthusiastic, which sometimes gives the impression of hypocrisy and insincerity.

Often offers simple ideas, briefly explaining them with examples from personal practice. He does not shy away from the exchange of opinions, arranges a collective discussion of important issues. Serious debate is considered boring. He prefers real, practically realizable proposals to theoretical and philosophical lengthy reasoning. Being in a tense state, gives the impression of a bored person who is not interested in the issues under discussion.

Most of the successful politicians and businessmen, artists and singers, managers and producers have taken place in the profession thanks to the use of sober calculation. They do not tend to stray from the intended path, being distracted by sentimental thoughts and wasting energy on emotional actions. In life, they are guided only by cold calculation.

Public opinion

It is not uncommon to hear negative reviews about successful people.

The following features of pragmatists cause outrage:

1. Cynicism. The belief that everything has a price in monetary terms, you can do any action to achieve positive results causes rejection. As a result, others consider them immoral.
2. Lack of authority. For pragmatists who seek profit in everything, only their own interests are important. They can listen to someone else's opinion, but they will take it into account only if it is in their interests. In other cases, they will not rely on other people's words, authority and actions.
3. selfishness. All efforts are applied only to achieve the goal. On the way to her, other people's emotions, losses will not stop him. The interests of others are not of interest, since the main thing in life is the result at any cost.

It is these qualities that cause a negative attitude that are necessary for the implementation of the plan. These people do not stop at obstacles, difficulties only temper their character. All this allows you to bring the work you have started to the end.

Conclusion

Anyone can develop the best features of pragmatism. To do this, you need to set specific goals, plan for the future, bring what you started to the end, not succumb to difficulties. There are not so many people who can be called pure pragmatists. In most cases, different abilities, inclinations and desires are present in varying degrees in one person.

Modern conditions require people to be able to plan, adapt to the rapid pace of life, and quickly respond to changing circumstances. A practical approach allows you to succeed, so we can say that a pragmatist is a person who is purposeful, and feelings and emotions do not really matter to him.

They are often disliked, envious of their assertiveness and energy. As a rule, ill-wishers are weak-willed, weak-willed individuals. Do you consider yourself a pragmatist or their critics?

Computer science as a technical science must, by definition, include pragmatics. This circumstance is to a certain extent misunderstood, and therefore the pragmatic side of informatics is much less talked about than its syntax and semantics. In fact, computer science is thoroughly permeated with pragmatics.

Charles Pierce defined the maxim of pragmatism as follows: “It is necessary to consider all the consequences dictated by a certain concept that the subject of this concept will have. Moreover, those that, according to the same concept, are capable of having practical meaning.

It is well formulated, but the fundamental concepts of pragmatics, which are values. In the technical sciences, instead of values, people often talk about criteria or norms. But in any case, we are talking about concepts that express people's preferences. In accordance with them, people set some goals for themselves, improving them. In the pragmatic sciences, the central place is occupied by value-target explanation, practice is understood as purposeful activity. Recall that the quantitative measure of any value is grade. It is the result of a corresponding process measurements. In this regard, it is appropriate to single out the main provisions of the theory of assessment measurement.

• 2. Measuring grades involves assigning some numerical or linguistic variables to them2.
• 3. Evaluation always has some dimension, which is determined by the nature of the value. Anonymous scores have no scientific meaning. There are points for knowledge, skills, beauty, honesty, and so on.
• 4. Rating scales are not universal, as evidenced by the widespread use of order scales, direct and proportional ratings, equal and half intervals, correlative pairwise ratings, etc.
• 5. The types of measurements also vary. Distinguish direct, indirect, cumulative and joint measurements. In indirect measurements, the value of a quantity is determined on the basis of a known relationship between the desired quantity and the quantities whose values ​​are found by direct measurements. Cumulative measurements involve a combination of repeated measurements. Joint measurements of oppositely named quantities are designed to provide a functional relationship between them, i.e. certain laws.
• 6. Understanding the process of measuring grades is not independent of the content of the theory, moreover, it is entirely determined by it.

Both in science and in the philosophy of science, the position of objectivists is very strong, who think that measurement is a simple act of fixing what is. A certain standard is chosen, with which the measured attribute is then correlated - that's the whole theory. There is no need to go into the intricacies of how theories are constructed. This opinion has been repeatedly refuted, but the objectivists do not let up. In order not to be unfounded, we will refer to an illustrative example. Not every economist understands that the price of a certain product is not a fixed value, but essentially depends on the chosen optimization parameter. If this is the volume of sales, then the price will be different than when maximizing the profit attributable to the capital advanced. The general conclusion is that what exactly should be measured and how, is determined based on the content of the theory.

7. Values ​​that don't generate goals are lifeless. Given this circumstance, it becomes clear that evaluation is a simultaneous characteristic of both value and the goal it generates. Estimates are not autonomous from the goals, they are bound to be closed on them, and therefore, they must be formulated taking into account the certainty of the goals.

Especially many collisions are connected with the understanding of the content of this provision. It is usually not taken into account at all. It is believed that the goal setting, which varies from one person to another, is incompatible with the solidity of science, which is determined by the foundation that is unchanged for all time. But such an opinion cannot be reconciled with the actual status of the pragmatic sciences. In each of them, the matter is not limited to immutable principles, but is realized only in setting goals, on which the value of the measured parameters depends. It is the goal that turns out to be the frame of reference, relative to which the parameters have certain values. The higher the significance of the parameter, the greater its value.

Software quality as a value.

So far, pragmatics has been characterized in the most general terms. Let us now turn directly to computer science, considering the quality software (ON). The very formulation of the question of software quality testifies to its pragmatic content.

Strictly speaking, software has been evaluated in one way or another since its inception, but only since the late 1960s. this process began to take systematic forms. In the second half of the 1970s. the first monographs appear1, the number of which then steadily increases exponentially. A multifaceted and persistent search is underway quantitative measures features of software, which is the subject of numerous works in the field of software metrics. Among such measures, two leaders are quickly emerging: first, the number of lines of code and, second, the number of errors per thousand lines of code. Of course, software evaluation is not limited to them. In table. 4.4 provides software quality factors according to GOST 28195-892.

Table 4.4. Software quality

	Software quality factor	Characterized property
	Reliability	It characterizes the ability of software in specific areas of application to perform specified functions in accordance with program documents in the event of deviations in the operating environment caused by hardware failures, errors in input data, maintenance errors and other destabilizing effects.
	Maintainability	Describes technological aspects that make it easy to eliminate errors in software and policy documents and keep software up to date
	Ease of use	It characterizes the properties of the software that contribute to the rapid development, application and operation of the software with minimal labor costs, taking into account the nature of the tasks being solved and the requirements for the qualification of maintenance personnel
	Efficiency	Characterizes the degree of satisfaction of the user's need for data processing, taking into account economic, computing and human resources
	Versatility	Characterizes the adaptability of the software to new functional requirements arising from changes in the scope or other operating conditions
	Correctness	Characterizes the degree of compliance of the software with the requirements established in the terms of reference, data processing requirements and general system requirements

Software metrics. The metrics of Halsted, McCabe, and Chapin are especially popular. Their indicators, as well as many others, allow not only to assess the complexity of the implementation of individual elements of a software project, to adjust the overall indicators for assessing the duration and cost of the project, but also to assess the risks associated with the project and make the necessary management decisions. At the same time, we have to state that the entire software quality program is replete with numerous problematic aspects. English researchers N. Fenten and M. Niel wrote about them very brightly in the article "Software Metrics: Achievements, Failures and New Directions".

The authors attributed the development of various metrics and their at least partial use not only by large, but also by medium-sized companies to the achievements. However, this gratifying process is met with significant difficulties. Academic activity in the field of creating software metrics does not find its continuation in the industry, in particular because many scientific developments do not take into account its needs. The metrics used in the industry are poorly motivated and not used effectively enough. In the course, mostly simple indicators that are easy to remember. More complex indicators are more likely to be exotic.

H. Fenten and M. Niel came to the conclusion that one should be extremely attentive to the requests of practitioners who prioritize software defects. When characterizing a particular software module, it is necessary to indicate the number of defects corresponding to its three levels - the lowest, middle and highest, determining the probability of a possible failure.

American researchers S. Kaner and W. Bond subject each of the proposed indicators to a thorough pragmatic analysis, including the one that was so widely promoted by T. Demarco, stating: "I know only one indicator that deserves to be put together once and for all: counting defects". In this regard, it is necessary to give a quite confident answer to 10 questions:

• 1. What is the purpose of this measurement?
• 2. What are the boundaries of this dimension?
• 3. What attribute are we trying to measure?
• 4. What is the natural scale for measuring this trait?
• 5. What is the natural variability of a trait?
• 6. What is the metric (function that determines the value of a feature)? What tool is used to carry out the measurement process?
• 7. What is the natural scale for the metric in question?
• 8. What is the natural scale of the instrument being read?
• 9. How do the values ​​of the attribute and the metric correlate?
• 10. What are the natural and predictable side effects of the tool being used?

All these questions are aimed at clarifying the real, and not the illusory pragmatic content of a particular feature. But since, contrary to Demarco, things never end with the selection of just one indicator, it is necessary to find a balanced indicator. The corresponding concept is theory of balanced system of performance indicators, was first developed in the field of management by R. Kaplan and D. Norton. From management, i.e. theory of management of organizations, it was rightfully transferred to computer science by A. Abran and L. Bouillone. S. Kaner and W. Bond became co-authors of this transfer.

• 1. Values ​​are always part of theoretical constructions. Their assessment is largely determined by the set goal.
• 2. The final criterion of effectiveness is the success of the functioning of the theory. In case of failure, the theory is revised. This is the mechanism of the growth of scientific knowledge in computer science.

Send your good work in the knowledge base is simple. Use the form below

Students, graduate students, young scientists who use the knowledge base in their studies and work will be very grateful to you.

Similar Documents

Definition of trademarks as designations that serve to individualize the products of legal entities or individual entrepreneurs. Classification of trademarks: figurative, verbal, combined, sound, holograms and three-dimensional.

presentation, added 03/20/2012


The role of trademarks and their varieties. The specifics of the sale of branded goods. Legal aspects of the use of trademarks. The practice of image advertising in the Russian market. Exclusive right to a trademark. Trademark as an element of advertising.

control work, added 02/14/2010


Essence, purpose, types of trademarks. Overview of the features of the use of a trademark. Organization of international licensed trade. Trade in counterfeit products. Measures taken when goods are identified that have signs of counterfeit.

term paper, added 11/28/2014


Corporate style: concept, functions, main elements. Trademark history. The value of cultural traditions in the design of trademarks. The phenomenon of erasing national and state borders. Search for figurative and semantic ideas in design creativity.

term paper, added 04/04/2018


Types and functions of trademarks. Basic requirements for a trademark, as well as rules of use. Legal protection of trademarks in the Republic of Belarus. The role of the brand and branding. Factors determining the level and dynamics of world prices.

abstract, added 07/21/2013


Involvement and types of games in advertising. Advertising as a means of communication. Sports Advertising Media and Sales Support: Criteria for Selecting Objects and Persons. The importance of advertising for the development of trademarks in the sports business. Processes of transformation in Russian sports.

abstract, added 03/23/2014


Trademark as an element involved in trade relations. Varieties and features of a trademark, its psychological impact on consumer preference and choice. The history of the Apple trademark. Principles of creation of commodity brands.

abstract, added 11/24/2011


Marketing planning as an integral part of a business plan in Russia. The main goals of the marketing program, directions for evaluating its effectiveness. The problem of development, formation and promotion of trademarks. Benefits and functions of trademarks.

test, added 02/28/2012

The problem of defining the boundaries of the subject area of ​​any scientific direction is primarily relevant in solving terminological and terminographic problems, including the tasks of term inventory, as well as in teaching the relevant disciplines.

Determining the boundaries of the subject area of ​​such a scientific direction as "Computer Linguistics" is one of the most difficult tasks, in our opinion. Typically, the boundaries of the subject area are established by compiling a list of headings and subheadings (directions) that form it.

The main difficulty in this case lies in the fact that computational linguistics is a relatively young science that originated at the end of the 20th century. This direction began to be actively developed abroad in the 60-70s, and it was primarily understood as the use of static methods in linguistics, hence the name "Computational Linguistics" (i.e. "Computational Linguistics"). In Russia, the related term "Mathematical Linguistics" became widespread in the 70s. In connection with the development of computer technologies and their active applications in linguistic tasks, this term as the name of science has been transformed, and science has received a clearer definition of "computer linguistics". Thus, we can say that there are two approaches in determining the directions considered under this term - this is our Russian approach and the foreign one.

As for the view of foreign linguists on the subject area of ​​computational linguistics, it can be noted that the Association for Computational Linguistics, which has regional structures in several countries of the world, is doing a lot of organizational and scientific work. The official website of this organization gives a general definition - "computational linguistics is the scientific study of language from a computational perspective. Computational linguists are interested in providing computational models of various kinds of linguistic phenomena". This organization organizes international conferences on computer Corpus Linguistics: Investigating language and COLING. Computational Linguistics is published quarterly in the USA. Relevant issues are usually also widely represented at various conferences on artificial intelligence.

From the point of view of the Western approach, the main direction of computational linguistics is Natural Language Processing (Automatic processing of natural language and speech). When analyzing documents (conference archives, content of basic sites) of the COLING Association for Computational Linguistics, it was noted that Western linguists include the following applied areas in the field of computational linguistics:

Computational Morphology and Syntax (Computer morphology and syntax).

• NLP (Automatic Language and Speech Processing).
• Digital Libraries (Digital Libraries).
• Information Extraction.
• Information Retrieval.
• Knowledge Representation and Semantics (Knowledge Representation and Semantics).
• Machine Translation.
• Speech Processing (Speech recognition and synthesis).
• Statistical Language Processing.
• Summarization (Summarization and annotation).

From the point of view of the Russian perception of the problem area under consideration, the main work in this direction is carried out by the Russian Association of Computational Linguistics COLINT, on the website of which you can find all the scientific reports presented at the conference on problems of computational linguistics. Although this site does not provide headings for problematic tasks, you can see that Russian linguists prioritize such areas as:

• Machine translate;
• Search and classification systems;
• Computer lexicography;
• Linguistic computer semantics;
• Corpus linguistics;

Formal models of analysis and recognition of language structures.

An analysis of existing textbooks and reference books does not yet give a complete and clear picture of this practical area of ​​linguistics.

Big Encyclopedic Dictionary: Linguistics, edited by V.N. Yartseva. does not include this term in the dictionary at all.

Famous Russian linguist Marchuk Yu.N. First of all, he defines computational linguistics as "the linguistic foundations of computer science", which actually involves solving problems related to the development and use of artificial languages ​​that provide communication between a person and a computer. But at the same time in his work "Fundamentals of Computational Linguistics" Marchuk Yu.N. consistently considers computer modeling of natural language, namely, morphology, syntax, representation of semantics and pragmatics in computer environments. In addition, the paper mentions such applied tasks as the organization of machine dictionaries, terminological data banks, and even discusses the basics of terminology.

According to the Russian linguist, professor of Moscow State University Baranov A.N. The term "computational linguistics" refers to a wide area of ​​using computer tools - programs, computer technologies for organizing and processing data - to model the functioning of a language in certain conditions, situations, problem areas, etc., as well as the scope of computer language models in not only linguistics, but also in related disciplines". In his work, Baranov A.N. identifies some areas of computational linguistics as basic - these are communication modeling, plot structure modeling, hypertext technologies for text representation, computer lexicography, machine translation, processing systems natural language.

The problem of defining the boundaries of the subject area also faces the developers of curricula for academic disciplines in the field of applied and computational linguistics. For these purposes, curricula in computational linguistics of such Russian universities as Moscow State University, Moscow State Linguistic University, Russian State University were analyzed.

The developers of the MSLU course program focus on computer modeling of natural language in solving artificial intelligence problems, the fundamental principles of language modeling, i.e. directions that are associated with modeling human-computer communication.

The developers of the course program at Moscow State University identify such tasks and areas as the problems of linguistic support for modern automated information systems, automatic processing of natural language, and the creation of dictionary and word processors.

The main areas covered in the course of computational linguistics at the Russian State University are: information retrieval, machine translation, terminology, terminology, terminography, computer lexicography, speech recognition and synthesis, computer-assisted language learning problems. The program of this university is closest to the program of the Ulyanovsk State Technical University.

The above analysis shows that in practice, almost everything related to the use of computers in linguistics is often referred to as computational linguistics, which is why problems are confused with practical solutions. Thus, when defining the boundaries of the subject area of ​​computational linguistics, it is necessary to more clearly distinguish 2 points of view:

1. AUTOMATIC LANGUAGE PROCESSING (Language Processing), which will include the tasks of analyzing and modeling the language structure, namely:

• graphematic/phonemic analysis of the language;
• morphological analysis;
• lexical and grammatical analysis of the language;
• parsing, or parsing;
• analysis and modeling of the semantic structure;
• the task of synthesizing language elements, incl. text generation;
• automatic linguistic statistics.

2. APPLIED DIRECTIONS OF COMPUTER LINGUISTICS, namely:

• Machine translate;
• speech recognition and synthesis;
• development and use of artificial languages, including programming languages, information systems languages;
• computer lexicography and terminography;
• linguistic foundations of information retrieval;
• automatic indexing, abstracting and classification of texts;
• automatic content analysis and authorization of texts;
• hypertext technologies for text presentation;
• corpus linguistics;
• computer linguodidactics.

This distinction does not claim to be complete, but it gives a more definite picture of the subject area of ​​this complex science, and can be used as a basis for developing a terminological dictionary or work program for the course "Computer Linguistics".

LITERATURE

1. Baranov A.N. Introduction to applied linguistics. - M.: Editorial URSS, 2001. - 360 p.
2. Grinev S.V. Introduction to terminological lexicography. - M., 1986.-106s.
3. Marchuk Yu.N. Fundamentals of Computational Linguistics: Textbook. - M., 1999. - 225 p.
4. Sosnina E.P. Introduction to Applied Linguistics: Textbook. - Ulyanovsk: UlGTU, 2000. - 46 p.
5. Yartseva V.N. Linguistics. Big encyclopedic dictionary. - 2nd ed. - Y41 M.: Great Russian Encyclopedia, 1998. - 685 p.
6. http://www.aclweb.org
7. http://www.dialog-21.ru

AT. With. Shchepin

CMA Small Systems AB

[email protected]

Keywords: artificial intelligence, environment model, fractal object, fractal memory, Turing machine, research structure, dialogue, information exchange

The aim of the work is to structure the pragmatics of dialogue interaction. The stages of perception by the intellect of information about the external environment, the construction of a model of the environment and the exchange of information about the external environment with other intellects are considered. The structures of the environment and memory of the intellect are fractal, that is, they allow infinite refinement and expansion at any point. The work is aimed at achieving a qualitative understanding of what knowledge is, what it can be used for and where the border between knowledge and ignorance lies. Dialogue is understood as an exchange of information models formed in the memory of subjects in the process of studying the external environment. The proposed model can also be considered as a model of interaction between a programmer and a computer. The results of the work may be of practical importance for the design of advanced structures of computer memory and software, including human-machine interaction languages.

1. Introduction

The scheme of the proposed model is as follows. Two intellects explore the environment, compiling its models, and then exchange the results of their activities in a dialogue. As a result, the total model of the environment formed by them collectively is formed in the memory of each intellect. In other words, there is an "exchange of experience", and the total experience can then be used by each intellect individually.

2. Environment as a fractal object

In the process of research or interaction of the subject with the environment, the environment appears in the form of interconnected objects of a lower level, which in turn can be subjected to further decomposition, and such a decomposition process is endless.

In other words, we imagine the environment as a fractal object in which any part is similar to the whole in the process of cognition - the representation of the object in the form of interconnected parts that are also objects.

The environment is structured in the process of cognition, it is not constructive to talk about the a priori structure of an unknown part of the environment - it is unknown and can be any. Moreover, any local part of the environment can be studied only partially; the environment always allows further detailing of the structure at any point. We know nothing a priori about the environment, except that it is fractal.

3. Decomposition of the environment into objects

3.1. Environmental research

In the process of research, the environment is structured and information about it must be recorded in the subject's memory in order to be used for purposeful behavior and be the subject of exchange in the process of communicating with another subject in a dialogue mode.

The goal is the structure of the representation of information about the environment in the memory of the intellect, for which it is necessary to analyze the process of studying the environment by the intellect. The research process can be broken down into the following steps:

• Scanning of the perceived part of the environment and selection of objects;
• Establishing links between objects;
• Entering into memory the structure of objects and their relationships;
• Change of "standing point" of the subject;
• Repeat previous steps.

3.2. Objects as anchor points in the environment

The selection of objects is the first stage in the structuring of the environment. Objects are determined by local sets of characteristic features that allow them to be distinguished from the environment in the process of perception. An object is not necessarily a physical body. Sections of a report or message may well be considered as objects representing a decomposition of a topic.

The pragmatics of displaying objects in the memory of the intellect is reduced to the information required to call the action programs necessary for the search and recognition of objects. The information about the object stored in memory should be sufficient to detect this object in the environment. This may be a visual image of the object or some description in NL, which can be stored in the subject's memory.

3.3. Links between objects

Relationships between objects are constructed by the intellect in the process of structuring the environment. Links between objects are established through processes performed by the subject, such as "track the relative position", "see", "walk the distance from one object to another", "understand the logical relationships between sections of the text". In other words, the pragmatics of displaying links between objects is reduced to the actions of the subject to identify these links.

Revealing a connection is the action of the intellect to establish a relationship between objects. Links are described by sets of parameters, which are enough to run a "program" that allows the intellect to recognize a specific link between objects.

Connections can be stored in the memory of the intellect in the form of information that controls its perception organs, as well as in the form of fragments of texts in NL.

The purpose of this work is a structural model. We will assume that both the connections between objects and the objects themselves are recorded in units (cells) of the memory of the intellect. It is necessary to understand how the cells should be interconnected, what should be the memory structure of the intellect to ensure the possibility of building a model of the environment.

4. Formation of a structural model of the environment in the memory of the intellect

4.1. Bypassing objects and building a model in the memory of intelligence

The main assumption of this work regarding the functioning of the memory of the intellect is that a model of the environment is formed in the memory and a certain "cursor" (mind's eye) is always directed to the memory cell that corresponds to the object of the external environment, where the subject's attention is currently directed. In this case, either the initial entry of an object or relationship (connection) into memory, or a comparison of perception with information previously recorded in memory, occurs.

It is also assumed that the 2 intellects participating in the thought experiment are trained, and they already have predetermined programs for recognizing objects and relationships, according to which the environment is structured in the course of their activity. For both intelligences, the sets of objects and relationships between objects that they can see in the environment are the same.

The intellect carries out the process of scanning the environment in order to select objects and establish links between them. The simplest assumption is that this process follows a tree-like algorithm: the first object is selected, then the first connection of this object with the next one, and so on. The "visible horizon" can be stored in memory as a simple sequence of cells that store objects and links, or as a tree if it was not possible to bypass all visible objects by links without returning.

From the principle of parallel scanning of the model and the environment, the above algorithm for studying the environment will be written somewhat differently:

• Find an object, including the link to the previous object.
• If the object and/or connection is not in the model, include them in the model.
• Back to step 1

It can be assumed that the structure of visual-motor images and the structure of linguistic concepts (identifiers) denoting objects and relations between them are formed in parallel in memory.

4.2. Environmental model requirements

4.2.1. Path planning. The model must provide the ability to plan a path to reach a specific object already included in the model.

4.2.2. Situation tracking. The model must provide a solution to the problem "Orientation on the ground" or "Return to the starting point". When moving in the environment, it is necessary to monitor the situation in order to correlate the observed objects with their display in the model at any time and provide the possibility of returning to the starting point.

4.2.3. Expansion of the visible horizon. When traversing objects included in the model, repeating the scan from points near the objects may reveal new objects. Accordingly, the model must be expanded at these points. Model extensions are tied to previously known points, which allows you to maintain the integrity of the picture of the world.

In a structural model, a link of the "horizon extension" type should be generated from the object, leading to a new fragment of the model.

There is a problem to distinguish the new object from the previously included in the model, because from the "new horizon" you can see the objects of one or more previous horizons. Such objects should not be re-introduced into the model; references to the descriptions of objects included in the model earlier should be made at the appropriate points. In other words, it is necessary to establish the fact that the observed object is known if its representation is already present in the model.

4.2.4. Detailing of objects. Objects, when approaching them, can be detailed, a new structure can be distinguished in them, that is, new objects and connections in that essence, which from afar seemed to be one indivisible object.

In this case, in the model, a connection of the "detail" type should be generated from the object, leading to a submodel of a lower level.

4.2.5. Recognition of an unknown environment. Getting into an unknown environment, the AI ​​must maintain the integrity of the world model, either by comparing the perceived objects with those already present in the model and thereby determining the standing point, or by building a new fragment of the model in memory, postponing for the future the attachment of this fragment to the previously built model.

4.3. Memory structure for the formation of the environment model

4.3.1. Requirements for memory cells. We will assume that the AI ​​memory cell can store information about objects and relationships between them. Both in the form of visual images and in the form of language constructions.

4.3.2. memory structure requirements. The process of interaction of AI with the environment described above leads to the need for AI to have a memory that is expandable at any point in it, so that the processes of "expanding the horizon" and "detailing" can be implemented.

This implies the need for a hierarchical structure of memory, a cell describing an object should, as it were, "open up" at a new level, and make it possible to write into memory a "new horizon" or a detail of the internal structure of the object.

At the same time, the structure of memory at all levels is the same, since objects and relationships between them are always remembered. It follows that memory must be fractal - the structure of any part and the whole are of the same type.

It is possible to imagine the AI ​​memory structure of each level as a kind of environment that allows you to form an arbitrary graph of objects and relationships. However, you can get by with a simpler tree structure.

The perception of an object or a connection between objects requires active actions - eye movements, for example. The perception of the entire visual scene can be imagined as a sequence of such actions that remember the sequence of objects, and then the construction of branches with the establishment of links to other objects from those already stored in memory. To implement such an algorithm, the potential possibility of constructing an arbitrary number of links from any object according to the principle of "horizon expansion" should be provided - here it will look like an expansion of the visual scene analysis area.

From this logic of perception follows the idea of ​​a branching tree-like structure of AI memory, where the first chain of objects and relationships is written into a sequence of cells, and then this sequence branches in an unpredictable way.

The same object or connection should not be entered into the model several times. Links to known objects are written in a special way. In a special cell referring to a known object, the chain of actions of moving the "cursor" along the constructed environment model is stored, leading to the object recorded there earlier.

4.3.3. On the model of the Fractal Turing Machine. Based on the requirements for the AI ​​memory structure formulated in the previous section, it is possible to propose a model of the Fractal Turing Machine, necessary and sufficient to implement these requirements in terms of structure. The essence of the model is that any cell of a regular MT tape can be detailed by a whole tape of the next level, and the cursor can move not only along the tapes, but also from level to level.

4.3.4. Arguments in favor of a tree-like fractal structure of memory. Listed below:

• The tree-like structure is the simplest option, sufficient for constructing a structural model for studying the environment.
• The structure of the neuron is also tree-like.
• AI behavior is easy to describe in the form of a tree of actions - situations.

• Well-structured computer programs are treelike.

5. Dialogue as a process of information exchange

5.1. Dialogue in terms of pragmatics

Two intellects exchange information in the process of dialogue. The initiative to conduct a dialogue may belong to one side or the other, and the parties in turn may have the initiative or be active.

Activity in the dialogue means the following. The active side actually explores the environment model built by the other side. This happens according to the same algorithm as the environmental exploration described above, due to the fact that the internal environment model generated by the AI ​​is also a fractal object. Asking questions, the participant in the dialogue reveals the objects and relationships displayed in the model of the other side. At the same time, the process of dialogue is structurally similar to the process of exploring the environment.

5.2. Stages of dialogue

Consider the stages of dialogue in the light of the analogy with the stages of research:

1. Identification of the known part of the environment model common to both subjects. In essence, this is the choice of the topic of the dialogue. Asking questions about the objects and relationships present in its model, the active subject finds a structurally matching set of objects and relationships in the model of another dialogue participant. This may be reminiscent of solving the Unknown Environment Recognition problem.
2. Search in the model of the passive subject for a fragment of the model of a part of the environment unknown to the active participant in the dialogue. The algorithm of such a search is reduced to bypassing the known objects of the matched part of the model and exploring the possibilities of detailing objects or the emergence of a “new horizon” on those objects that do not have such continuations in the model of the active participant in the dialogue. Research is carried out through appropriate questions.
3. "Copying" knowledge into the environment model of the active subject. The active participant can select the most interesting object containing new information, move the "cursor" in his model to this object, move the "cursor" of another dialogue participant to the same object and start completing the construction of his model in the dialogue mode, asking questions about new objects. and connections between them. This may ask questions such as:

• What objects are there?
• How is each object related to the origin?
• How are the objects related to each other?

6. Conclusions

1) The fundamental property of the intellect is the ability to accumulate information about the external environment and to organize, on the basis of this information, one's actions, one's behavior in the environment.

2) The structure of the memory of the intellect corresponds to the structure of the possibilities of studying the environment by the intellect. It should be fractal and most likely tree-like.

3) From the point of view of pragmatics, the process of exploring the environment by the subject and the study of the memory of another subject in the process of dialogue between subjects have a similar structure.

4) AI (computer) can be considered as one of the subjects of the dialogue. In this case, the structural construction of computer memory according to the principle of the Fractal Turing Machine can be very effective, as it corresponds to the structure of human memory.

5) Information technology paradigms can be based on artificial intelligence models, and not only on concepts such as object-oriented programming, ideas about automating document processing on the desktop, or on the ideas of creating virtual reality by simulating objective reality. In particular, navigation mechanisms are often used in software systems. By exploring abstract navigational models, AI, as a scientific discipline, can contribute to the practice of building software systems.

6) There is a practically useful implementation of the concepts of this work. This is a computer "notebook" or, one might say, a personal information system, which is an infinitely expandable tree structure of text lines. It turned out to be convenient for writing texts by gradually building and then detailing the structure of headings and subheadings (with their rather frequent restructuring). It turned out to be convenient for writing programs that immediately turned out to be well structured. At the same time, the implementation, like the concept, is very simple, which was the reason for the birth of this message.

Structured model of dialog "s pragmatics based on fractal

intelligence-environment model

V. S. Shchepin

key words: artificial intelligence, environment model, fractal object, fractal memory, Turing machine, structure of investigation, dialog, information exchange.

The aim of this paper is to investigate the structure of pragmatics in dialog interactions. The following logical scheme was chosen: An Intelligence must first of all percept an environment then build an internal model of this environment and then be capable of exchanging information from the model with another Intelligence, natural or artificial. How could it be done? This is a problem. The first suggestion is that environment is fractal i.e. it "s decomposition may be infinite at any point. The second suggestion is that an Intelligence memory should be fractal too. Fractal memory is needed to reflect a fractal reality of an environment.

During his activity in the environment the Intelligence must structure the environment in his memory. A structured model of environment should include objects and object's relationships descriptions. The object's description must contain information sufficient for object's recognition by the Intelligence. The objects relationship's description must contain information needed for checking the relationship in the environment. In fact relationships between environment’s objects are established by some actions of the Intelligence. In a process of the dialog between two Intelligences two structured environment models are first compared for common fragments and then extended by the unique fragments from both sides ensuring an exchange of experience. An environment is considered as static for simplicity.

The study of these questions may give us a qualitative understanding of what is knowledge, how it can be used and where is a boundary between known and unknown. The structured model proposed by this paper can serve for example as a model of interaction between a programmer and a computer. Hence the ideas expressed here may be useful for a computer memory and computer software perspective design including more powerful languages ​​and operating systems for man-computer and for computer-computer interactions.