Lecture
Это продолжение увлекательной статьи про словарь по искусственному интеллекту.
...
solved.
Real object, process, phenomenon, situation, etc. in the physical world, for which there is a special naming expression in some language. This naming expression is the designate for the given denotation.
Representation of the inference procedure in the form of a tree, the vertices of which are either the original formulas or formulas obtained in the process of inference.
The representation of the search process in the form of a tree, each vertex of which is associated with the value of the search key in such a way that all the smaller keys are concentrated in its left subtree, and all the big ones are in the right one.
Representation of the result of the parsing stage in linguistic processors in the form of a sentence parsing tree, at the vertices of which there are lexemes corresponding to the subject, predicate, addition, etc., and the arcs indicate the connection between the vertices in control. Used at the stages of deep syntactic and semantic analysis of the sentence.
A structure consisting of decision nodes and alternatives corresponding to these nodes. Movement by D.R. can be carried out randomly or on the basis of local information about success, which is available in the nodes. As a result of a successful search on D.R. a path is formed that leads from the root of the tree the initial situation to that node of the tree that corresponds to the target situation. In the process of movement by D.R. Often there is a need to return to the previously passed nodes, which is carried out using the backtracking procedure.
Representation of the system of components, the root of which is the full component, and the hanging nodes are point components.
A special kind of tree in which one or several vertices correspond to the goals, and the remaining vertices are subgoals of these goals. Arcs show how to decompose targets in subgoals.
Special naming expression for denotations that exist in external to this system of the world. All values of the system of denotations are recorded in the form of knowledge about D. In some cases, D. is called a unique name, label, key.
The highlighted word (or combination), which serves as a marker for understanding the natural language texts of the system. This marker is included in the left part of the rules of inference, sequence or products. When a marker appears, the corresponding rule is triggered. D. is selected from a special dictionary and he is artificially attached to semantic uniqueness, which allows him to denote a class of (synonymous) concepts.
The most typical attribute value attributed to an object, if its value for this object is specified.
An expression of the form B 1 Ú B 2 Ú ... Ú B n ¬ A 1 & A 2 & ... & A m , where (&, Ú , ¬) are respectively symbols of conjunction, disjunction, and implication. D. reads like this: "If A 1 , and A 2 , and, A m , then B 1 , or B 2 , or, ..., B n ". The right or left side of an implication may be empty. In this case, D. is interpreted differently. For D. of the form B 1 Ú B 2 Ú ... Ú B n, the interpretation consists in asserting the existence of a fact B 1 Ú B 2 Ú ... B n ; and for D. of the form ¬A 1 & A 2 & ... & A m, the interpretation consists in the statement that (A 1 & A 2 & ... & A m ). If in D. the left and right parts of the implication are empty, then it is called empty. D. is used in the method of inference based on the use of resolutions, as well as in the Prolog programming language.
Disjunct, which is either an element of the original set, or some clause that precedes the considered one in the derivation.
The empty set, arising from the exhaustion of the original set, clauses in case of successful completion of the inference process, based on the principle of resolution.
Disjunct containing no more than one positive letter.
Logical operation (bundle) for n> 2 expressions. The result expression is false only when the source expressions are false. To refer to D. the standard sign is используется (less often +).
Analysis of the text in terms of its communicative function and structure. The theory of discourse studies the laws governing the construction of texts, the compatibility of text units and text fragments.
Dissonance arising from the subject, when he has at the same time two conflicting knowledge about the same object, subject, situation or phenomenon. D.K. there is an incentive to start some activities to eliminate the contradiction in knowledge. In intelligent systems, D.K. used in knowledge bases to make knowledge active.
The conclusion of the statement in logical calculus, which provides an explicit construction of all elements involved in the output. Such is not, for example, a widely accepted way of proving "by contradiction" used in classical mathematics.
The logical following of this formula from this aggregate of previously derived formulas.
The aggregate of values of a certain information item placed in the database. D. is determined by its attribute.
The method of controlling parallelly running asynchronous processes for solving problems in which information about the terminated processes and the results obtained is “posted” on the DO, to which all processes (and / or programmers) that are waiting for the desired results have independent access. BEFORE. often used in expert systems, intelligent robots and other intelligent systems.
One of the basic laws of reasoning characteristic of traditional formal systems. He argues that the expression (AV ¬A) is identically true. Z.I.T. was criticized and rejected by logicians who are in the position of intuitionistic mathematics and constructive mathematics.
One of the basic laws of reasoning characteristic of traditional formal systems. He claims that the equality A = A always holds.
A collection of information that forms a holistic description, corresponding to a certain level of awareness of the described issue, subject, problem, etc.
Knowledge that is stored in the memory of an intelligent system so that it is immediately available for use after accessing the corresponding memory field. In the form of ZD usually information is written about the properties of the subject area, the facts that exist in it, etc. information. According to the presentation form Z.D. contrasted with procedural knowledge.
The collection of information about the subject area stored in the knowledge base of an intelligent system. In Z.P.O. includes facts related to the subject area, patterns characteristic of it, hypotheses about possible connections between phenomena, processes and facts in it, procedures for solving typical problems in this problem area. Z.P.O. enters the knowledge base of knowledge engineer. In the process of functioning of the intellectual system Z.O.P. can be replenished. Z.P.O. used when searching for solutions to problems arising in expert and other intellectual systems.
1. Knowledge of how to solve problems in a given subject area.
2. In a natural language, knowledge about the pragmatic component of texts. (See also User model, Focus of attention).
Knowledge stored in the memory of the intellectual system in the form of descriptions of procedures by which they can be obtained. In the form of Z.P. information about the subject area describing the ways of solving problems in this area, as well as various instructions, techniques, etc. are usually described. information. According to the presentation form Z.P. opposed to declarative knowledge.
Knowledge accumulated by the intellectual system in the process of its functioning, as well as knowledge embedded in it a priori, but not having the status of absolute truth in this problem area. Often Z.E. associated with the reflection in the knowledge base of human (informal) experience in solving problems.
Knowledge available to a specialist in a particular subject area.
The constant assigned to an attribute in the database.
The value of a variable that is automatically assigned to it, if its value is not specified.
A set of models and methods for technical systems to carry out procedures characteristic of visual perception in living organisms. In the framework of Z.M. the tasks of selecting objects from the background, their identification, input into intermediate memory, transcoding into internal representations, etc. are solved. Systems Z.M. characteristic of intelligent robots and other intellectual systems.
Oriented graph with properties: 1). When exciting (transmitting information) input arcs leading to a certain vertex, either a conjunction (AND) or a disjunction (OR) is realized. In the first case, the vertex is excited (becomes active and receives information) only when all the arcs entering it are excited. In the second case, to excite a vertex, it is sufficient to excite any arc entering into it. 2). When the vertices are excited, either all arcs out of the vertices (AND) are excited, or only one selected by the vertex (excluding OR for the number of arguments equal to the number of outgoing arcs). Often, the AND / OR term is understood as a graph for which the first property is fulfilled, and for output arcs, the AND / OR data always occur. They are widely used in planning systems for the efficient behavior of autonomous robots and in other artificial intelligence systems.
The procedure for establishing the properties of interest to the researcher in the phenomenon or object under study. For intelligent systems, I. often means checking that a given system really solves the tasks for which it was created.
Characterization of knowledge required to solve the problem.
Obtaining information about the subject area from specialists and expressing it in the language of knowledge representation. OF. used when building an expert system or knowledge base.
Development of tool software for solving problems of artificial intelligence. In II-P. programming languages are being created that are focused on the characteristics of artificial intelligence tasks, intellectual aids, knowledge representation and manipulation languages, empty expert systems and shells, and other tools.
Communicative capabilities of a specific type of speech act. For example, the ability to express a question, a doubt, a request with an interrogative sentence such as "Would you not come?" .
One of the components of the speech act, along with the locus and perlocution. Performing actions through speech: motivation (request, order), question, doubt, approval, promise.
Reproduction of procedures for the formation of purposeful behavior of humans and animals in the outside world, depending on the situations arising in it. For I.I.P. special models and methods of planning activities are being developed. Imitation of intellectual behavior is widely used in intelligent robots.
Reproduction of the program through or with the help of special equipment of individual processes characteristic of human and animal thinking (recognition of situations, making decisions about one's behavior, understanding natural language texts, etc.). In the artificial intelligence, I.I.M., as a rule, assumes not the identity of the processes occurring in the brain and in the technical system, but the coincidence of the results of solving the same tasks.
Logical operation (bundle) for two expressions. The resulting expression is false when the first expression is true and the second is false (the operation I. is noncommutative). Standard implication designation: ®.
The method of transition from particular observations to a general pattern, which is satisfied by all particular observations.
Finding patterns that obey all known up to this point of observation. Найденные закономерности могут опровергаться новым наблюдениями.
Математическое доказательство справедливости некоторой закономерности, основанное на выдвижении гипотезы по конечному числу фактов и обоснований к изменению этой закономерности.
Специалист, основной задачей которого является проектирование баз знаний и наполнение их знаниями о проблемной области. В процессе этой деятельности И.П.З. выбирает формупредставления знаний, удобную для данной проблемной области, организует приобретение знаний из различных источников (официальные документы, учебники, монографии и т.п.), а также в результате общения с экспертами-специалистами в данной проблемной области.
Раздел искусственного интеллекта, в рамках которого решаются проблемы, связанные с извлечением знаний, приобретением знаний, представлением знаний и манипулированием знаниями. И.З. служит основой для создания экспертных систем и других интеллектуальных систем.
1. Научное направление, в рамках которого ставятся и решаются задачи аппаратного или программного моделирования тех видов человеческой деятельности, которые традиционно считаются интеллектуальными. (См. Представление знаний, Обучение, Общение, Объяснение).
2. Свойство интеллектуальных систем выполнять функции (творческие), которые традиционно считаются прерогативой человека.
Способ работы с экспертом при приобретении знаний, когда инженер по знаниям выступает в роли интервьюера.
In a broad sense - an explanation, an interpretation of something. In programming, the process of translating a program written in a high-level language into object code in such a way that the program is stored in computer memory in its original form, and translation into object code is carried out in parts, as necessary. In artificial intelligence - the establishment of communication between the two systems of descriptions, which allows you to understand one system at the level of another.
A set of technical and / or software tools that provides the interface of two or more elements of the system for their joint operation in this system. A typical example of technical I. is a set of design parameters of telephone sets and telephone channels, allowing to connect any telephone set to any telephone channel.
The set of software and hardware that provides a common intelligent system with the user in a limited natural domain of the problem area. The composition I.YE.A. includes dictionaries that reflect the vocabulary and vocabulary of the language, as well as a linguistic processor that analyzes texts (morphological, syntactic, semantic, and pragmatic) and synthesizes answers to the user.
The interface, which includes the means that allow a person to conduct common with a computer, without using special programs to enter into a computer.
A hypothetical device that can replace the human brain and (or) realize all the functional properties known about the brain. In artificial intelligence under I.M. understand the repetition of artificial properties of the properties inherent in the brain.
Text, (instruction, monograph, photo, film, etc.), observation or a specialist professional reporting the necessary information. From I.Z. information is collected, transformed by the statements recorded in the memory of the intellectual system.
The formal system defined by the quadruple , where T is the set of basic calculus symbols; B - syntax rules, with the help of which arbitrary elements are generated from T elements; A is the set of a priori true elements of calculus (axiom of calculus); Р - a set of semantic rules (rules of inference), with the help of which some elements of the system generate others.
See Calculus Propositional.
Calculus, in which axioms are given in the form of sequences.
The object of study in mathematical logic, which is based on the concept of a formal system. Artificial intelligence uses various ILs: predicate calculus, propositional calculus, calculus of classes, calculus of relations, multi-sorted and multi-valued logic, etc.
Calculus, in which, along with formulas for calculating statements, formulas are used which may include relations (predicates) connecting the groups of elements of the calculus and the quantifiers of community and existence.
Predicate calculus, in which under the quantifier sign there can be no predicate symbols. (See also the community quantifier, the quantifier of existence.)
A formal system, the basic elements of which are statements - undifferentiated sentences, for which at any given moment it can be argued that they are either absolutely true or absolutely false. I.P. studies the connections between these statements, which are defined by logical connectives: negation, disjunction, conjunction, implication, etc. is an axiomatic system, and for classical I.P. All axioms are identically true statements, and the rules of inference do not change this property. With the help of I.P. all identically true statements are generated, and only they.
Predicate calculus, in which all or some of the predicates are labeled to bind them to certain situations. Each situation is defined by a description in which non-title expressions are involved, and those that are associated with a given situation. As axioms I.S. the usual axioms of situations and characteristics of these situations are used in the problem area for which I.S. is used. (See also Situational Management.)
A method of describing a fragment of space known to a subject with placeholders in it. There are a number of KK variants that differ from each other in complexity and detail, for example, a map-survey and a map-path. With the help of K.K. It studies how a person perceives spatial situations and displays them in his memory. In intelligent sistemK.K. used to display spatial situations in knowledge bases and when working with professional experts, when a knowledge engineer receives information from them related to spatial situations.
Establishing the connection of phenomena or facts. In a strict form, K. establishes causal relationships between phenomena or facts. In a broader sense, K. establishes the influence of some phenomena or facts on others. In this broader sense, K. reflects in knowledge models in the form of causal networks and scenarios. With a narrow understanding of K. in the same models leads to causal networks.
In the narrow sense, it is a pointer to the truth area of a statement. Examples of K. in this sense can be lexemes: "always", "almost never", "for many", "in about half the cases", etc. In formal systems, as a rule, two quantifiers are used, called the quantifier of generality and the quantifier of existence. The first one corresponds to the lexeme "always" and "for all", and the second one - "exists". In a broad sense, K. can mean any value of a linguistic variable (for example, "many", "often", "far", etc.). It is in this sense that K are used in pseudo-physical logic and in situational control.
Attributing evaluations (including numerical) to expressions of the formal system. These estimates are sometimes called quantifiers. Estimates can characterize the degree of likelihood of expressions, priority in solving a problem, etc.
A special pointer to the fact that some containing variables applies to all formulas obtained by substituting, instead of the variables listed in this index, any values from the domains of these variables. K.O. denotes how where xi are the names of those variables to which its action applies (related variables).
A special pointer to the fact that some P occurs (or is true) with some variables listed in this pointer, and specific values that ensure this are not indicated, but only that they exist is recorded. Variables listed in the index are called bound. Standard K.S. denoted by where x i are variable names that are related.
The introduction of relations on the set of objects or phenomena that allow them to be divided into classes with the establishment of the genus-species, element class, integral part, etc. relations between classes. See also Taxonomy, Clustering.
The method of splitting objects or phenomena into classes based on some proximity relationship in the feature space. See also Taxonomy, Classification.
See Disjunct.
The section of philosophy that studies the problems associated with obtaining and using human in the process of activity.
A complex of scientific disciplines (cognitive psychology, the theory of argumentation, etc.), united by a single subject of research - a reflection in the cognitive structures of a person surrounding reality and the study of the mechanisms of reasoning about this reality.
See Declarative Knowledge.
The operation of attributing one element to another in such a way that a new arbitrary element is obtained. With the help of K., for example, words of a language are formed from letters, from words, punctuation marks and a space character - a sentence.
A unipolar or bipolar attribute, a pair of alternatives, a parameter, a scale, or a pair of opposing relations of the individual to the object or any side of it. K.L. used in the fashion of repertory grids developed by Kelly and his followers to identify those subjective ideas that guide people in their professional and everyday activities. These approaches to the identification of subjective knowledge are used in knowledge engineering in acquiring knowledge from professionals to fill in the knowledge bases of expert systems.
See Concept.
Logical operation (bundle) for n> 2 expressions. The resulting expression (conjunction of source expressions) is true only when all source expressions are true. To denote K. the standard sign is & (less often), as well as the multiplication sign in the form of a point. In many cases, the sign K. may be omitted.
See computer linguistics.
The linguistics section, whose task is to study problems related to computer text processing: the organization of a natural language interface, machine translation and referencing, statistical analysis of dictionaries and texts on computers, automatic speech recognition.
The unit of measurement of the output machine's performance (from the English. Logical Interence PerSecond), equal to the number of logical inferences made in one second. As a rule, from 10 to 100 computer commands are required to implement a single inference.
Any constant, variable, or negation.
The science of the right ways of reasoning. In the classic version consists of the theory of concepts, the theory of judgment and the theory of reasoning. For a long time Aristotle's teaching on syllogistic inference was associated with L. Syllogistic was the first deductive system that emerged in science. The basis of L. is the concept of an axiomatic system. The power of pure logic, distracted from the semantics of the domain, lies in the generality of its methods and provisions. It is important to note that L. is the science of thinking in terms of concepts, and not of knowing the world through thinking of concepts. This shows that in intellectual systems, purely logical problem solvers cannot exhaust the entire supply of funds necessary for reconstructing intellectual activity. Based on L. at the end of the XIX century. A mathematical logic began to be created, based on set-theoretic categories and the concept of a formal system.
A logic in which formulas are evaluated by values interpreted as the probabilities that a given formula takes the value "True." With withdrawal rules in L.V. procedures are associated that allow the probability estimate of the truth of the derived formula to be calculated from the well-known truth estimates for the assumptive formulas.
A kind of epistimistic logic in which all statements are supplied with quantifiers that evaluate the degree of plausibility of these statements.
Relationship logic, in which relations (predicates) or special operators characterize time dependencies ("earlier", "will be", "simultaneously", etc.). Another type of L.V. are the so-called logic, in which one of the arguments of the predicate is time (state, situation).
A formal system in which it is assumed that the quantifiers of community and existence can connect not only individual variables, but also predicate or other functional symbols.
The logic in which only two values are considered as truth values of expressions: 0 and 1, interpreted as absolute false and absolute true.
The system of reasoning about the patterns of action in a certain problem environment. L.D. relies on temporal logic and spatial logic, as well as on the properties of a particular environment. L.D. used in intellectual works and expert systems. For LD characterized by non-monotonous conclusions.
The generic name for the logic of norms, describing the normative predicted behavior, and the logic of estimation, which describes the evaluation characteristics for various statements. Used when organizing the behavior of intelligent systems.
The system of reasoning, explicitly taking into account the dynamics of the objects to which these arguments are attached. If time is explicitly included in the argument, then LD coincides with one of the variants of the time logic. If the dynamics is given by the laws of changing situations, then LD turns into situational calculus. L.D. It is used to model the functioning of open systems, in particular, open databases and knowledge bases, as well as in all intelligent systems dealing with a dynamic model of the world.
A collection of reasoning that goes around in everyday life and reflects the system of values, motives of actions and goals of people. In intelligent systems L.Z.S. It is used in cases when, when reproducing the activities of a professional expert, it is not possible to build a formal system into which the expert’s reasoning procedures could be loaded.
A formal system that describes the rules for the formation of general statements based on a finite set of particular statements. In L.I. all assertions are weighted by plausibility estimates characterizing the truth of these assertions.
The logic used in formal systems that are not based on classical constructions that go back to set theory. and on speculative constructions. In the reasoning about these constructions, the law of the removal of double negation and the law of the excluded middle turn out to be inapplicable. L.I. It is widely used in proving computer theorems and in intelligent intellectual systems.
The logic in which the relationship characterizes the types of relationships that coincide with the cause-effect or close to them in content.
The logic in which various imperatives are used as operators. Close to the logic of action. Used in intelligent robots and other intelligent systems.
A logic in which only constructive proofs are allowed. () L.K. underlie constructive mathematics, closely related to computability problems on computers and other devices with limited memory.
Logic based not on the content side of statements, but on syntactic categories and their structural (operational) links. The basis LM lies the concept of a formal system. Different interpretations of the formal system lead to different logical calculi. The most famous of which are the propositional calculus (propositional calculus), predicate calculus, situational calculus, multi-valued logic, etc.
The logic in which the natural numbers 0,1, ..., k are the values of the truth of variables.
The logic of a closed world, equivalent to some formal system. In LM the principle of monotony operates: if an assertion is obtained at some output step, then its truth at subsequent output steps cannot be changed.
The logic of the open world. In L.N. violated the basic principle of monotonous logic. If an assertion is obtained at some output step, then when a new information (new facts) arrives in the system, the truth of this conclusion may disappear. L.N. characteristic of most intelligent systems dealing with complex subject areas, for which it is not possible to obtain a priori exhaustive closed description.
The logic in which odd quantifiers are used is often fuzzy linguistic variable “frequency” quantifiers: “almost never, almost always.” Reasoning with such quantifiers requires special techniques for finding the quantifier, which must be attributed to the conclusion when the package is marked with certain quantifiers.
See Logic deontic.
See Logic deontic.
A formal system in which the quantifiers of community and existence can only bind individual variables, but cannot bind predicate symbols or other functional symbols.
The logic characteristic of the propositional calculus.
A formal system in which the axioms are used, which are characteristic for describing the possible locations of objects in a three-dimensional (or two-dimensional) space, the distances between them or pieces. L.P. allow to reason about the spatial location and relationship of objects for the case of absolute and relative coordinate systems and for the case when variables such as distance, local size or characteristics of the relative position of objects are specified as linguistic variables. In L.P. allocate the logic of distances and the logic of the relative position of objects in the metric and topological (blurred) versions.
Logic reflecting the perception by the subject or the artificial system of laws of the external physical environment. The special feature of L.P. is the presence of blurry scales on which objects are projected with which the logic deals. Examples of L.P. are temporary logic, spatial logic, action logic, etc.
See logic fuzzy.
Reasonings, in which, in the absence of explicit information necessary to continue the reasoning, the intellectual system or person refers to his memory and uses the information contained in it, intended for those cases when the necessary information is missing. The introduction of the default mechanism leads to the fact that L.R.U. becomes a monotone logic. L.R.U. widely used in open databases and knowledge bases.
A formal system that uses operators of the type "knows", "wants", "believes", etc.
A limited part of the space in which a certain object is completely placed, whose external borders coincide with the boundaries of L. The concept of L. is used in spatial logic.
One of the components of a speech act is one's own speaking, characterized by diction, speed of speech, its correctness, etc., without taking into account the intentions of the speaker and the effect achieved. The other two components are illokution and perlocution.
The calculus, which uses the operation of functional abstraction (conversion) xM, defines a function whose values for any argument are the dummy of this argument instead of x in all its occurrences in M. Such calculus is widely used in formal database models.
Theoretical structure, which reflects all the formal aspects of the functioning of some real or hypothetical device. Examples MA can serve as a finite-state machine, a Post machine, a Turing machine, and many other models studied in mathematics, cybernetics, artificial intelligence, and other sciences.
The control unit database in information systems. Specialized processor with its own memory, performing request processing.
The control unit of the knowledge base in the fifth generation machine. A specialized processor (system of processors) that performs request processing and forming responses in a certain subject area based on the use of sets of facts and knowledge about the subject area, presented in the form of rules, as well as inference mechanisms.
An abstract machine (a set of software tools) with which a hypothetical computer is simulated for a user, possessing practically unlimited random access memory and an expandable set of instructions. Mv uses to simulate the final memory and the basic set of commands.
A specialized processor (system of processors) that implements in parallel the basic operations that are characteristic for knowledge-based knowledge.
Abstract machine consisting of endless into both sides of the tape, divided into cells, and the control head. Ribbon cells can be empty or marked with a special character. The control head moves along the cage. For one tact of MP executes one of six basic commands: shift the control head one cell to the left, a similar shift one cell to the right, inscribing the marking symbol in an empty cell, conditional transition and stop. From the sequence of such teams renumbered by the natural numbers, such programs form the functioning programs of MP. Before starting work MP it is necessary to fill the necessary cells of the tape with marking symbols and position the control head against a certain cell. After that, MP will execute com *** at program number one. If these are not commands for shifting or stopping work, then the next executable com *** and the programs after this one are determined by special pointers (references) included in each com *** at the record and erasing the marking characters. In a conditional transition command, the choice of a new command depends on whether the cell contains a void cell or is marked with a symbol. Depending on the situation, a transition occurs to the program specified in the conditional transition command. Mn according to the results of work, it is equivalent to the Turing machine, but its functioning is slower. Like the Turing machine, MP serves to refine the intuitive notion of an algorithm.
A computer consisting of tens and hundreds of parallel processors. Design M.S. allows any processor to communicate with any other processor like a telephone exchange. Speed MS reaches tens of billions of operations per second.
Abstract machine consisting of endless into one side of the tape, divided into cells, and a control head that can move along the tape. Characters of the input alphabet, including a blank character, can be placed on the tape one by one in a cage. The control head can be in one of a finite number of internal states, one of which is special. It corresponds to the shutdown of M.T. Each step of the work consists in the fact that the control head for the pair (the observed symbol in the tape cell, opposite which the control head is located; - internal state of the head) produces a triple (new cell contents - new internal head state - head shift one cell to the left or right or maintaining the position of the head). The work of M.T. ends when the control head enters the end of work state. The initial filling of the tape and the initial position of the control head, together with its initial state, are set from the outside. Actions M.T. at each step are determined by the final table, the size of which corresponds to the number of characters in the external alphabet and the number of internal states of the head. M.T. is a model of a universal computational process, since it is possible to build a universal MT, which will imitate the work of any particular MT. In this sense, the universal MT. can be considered as a mathematical model of a computer built according to traditional architecture. M.T. is one of the possible refinements of the concept known in discrete mathematics. Languages generated by the work of MT, are called recursively-listed.
See Architecture Streaming.
See. Intellect is artificial.
The method of organizing interfaces based on listing the alternatives and supporting the possibility of selecting the desired one using the cursor and / or explicit indication of its name.
Evaluation of the truth of an event or fact, the value of which is obtained by adding to the unit the value of the trust function.
Knowledge of the intellectual system of knowledge, which is stored in its knowledge base, or about the procedures that can be performed with those stored in the knowledge base. M.'s introduction is a recursive process. M. in natural language texts can be correspondingly with phrases like "I know that Ivanov does not know how to swim" or "Sidorov suggested that Petrov does not know algebra . "
Products to be included in the production system to indicate the order in which the products are included in the list of finished products.
Transferring the properties of one object (phenomenon) to another on the basis of a characteristic common to both objects being compared ("speaking of waves", "fermentation of minds")
Language for describing other languages. Most often, the metalanguage uses a notation, in which the own characters of the described language are the terminal symbols of the metalanguage.
A method for solving problems of integer linear programming and search problems on tree structures using heuristic rules for clipping search options based on local assessments of the feasibility of further search in a given direction, generated in the process of implementing the method.
In knowledge engineering, a technique by which knowledge is obtained from expert professionals. A knowledge engineer plays the role of a reporter interviewing. He asks questions whose purpose is to clarify the information given by the expert regarding the subject area in which the expert works. There are special techniques that are included in the standard MI, which makes the conversation focused and effective.
See search downward.
See Search ascending.
A set of inference rules and output control strategies (the application of these rules). Extreme case m.B. there may be an arbitrary application of the rules of inference, as is done in logical calculi.
Reception used in knowledge bases. It means that on the set of information units, classifying relations like "class-element", "genus-species", etc. are introduced. At the same time, information relating to all elements of a class or to all species of a genus is contained respectively in the description of a class or genus, and the information units subordinate to them inherit this information when necessary.
Архитектура вычислительной системы с несколькими одинаковыми или разными параллельно работающими процессорами, каждый из которых выполняет свои команды над своими данными.
Множество, характеристическая функция которого может принимать значения из отрезка [0,1]. Значение характеристической функции для некоторого элемента характеризует степень принадлежности этого элемента к множеству.
Объект (реальный, знакомый или воображаемый), отличный от исходного, но способный заменить его и в рамках решаемых задач.
Модель процесса решения задачи человеком, опирающаяся на процедуру установления сходства данной задачи (или составляющих ее подзадач) с задачами, решение которых уже известно.
Описание процедур решения задач в некоторой предметной области. В М.В. задается полная структура функциональных связей для элементов предметной области, связанных между собой соотношениями, позволяющими находить значения одних элементов через другие. Задание исходных целевых элементов приводит к поиску в М.В. путей, ведущих от исходных элементов к целевым. Если хотя бы один такой путь существует, то по нему строится программа решения поставленной задачи. М.В. обеспечивают автоматический синтез программ.
Модель, остающаяся неизменной при работе с ней. В процессе функционирования интеллектуальной системы М.З. в отличие от открытой модели нельзя добавлять новые факты и закономерности. Все утверждения, полученные в М.З. окончательны и абсолютны.
Описание знаний в базе знаний. Известны четыре типа М.З.: логические, в основе которых лежит формальная модель; сетевые, в основе которых лежит семантическая сеть; фреймовые, основанные на фреймах ; продукционные, основанные на продукциях. Каждая такая М.З. определяет форму представления знаний.
Гипотетическая модель, описывающая устройство когнитивной структуры (структуры знаний у человека). Для интеллектуальных систем М.К. совпадает с моделью знаний.
The model of the subject area from the list of all concepts used to describe this area, along with their properties and characteristics, classifications of these concepts by type, situation, signs in a given area and the laws of functioning processes occurring in it. M.K. built when immersing the description of the subject area in the knowledge base of the intellectual system.
One of the models of logical semantics used in artificial intelligence. The basis M.K. lies the idea of a set of possible worlds, each of which is defined by a formal system. The transition from one possible world to another in the framework of M.K. carried out with the help of a special relationship, the properties of which may vary.
The model in which the process of solving problems by man is explained by the analogy with the movement through the maze. The maze sites correspond to intermediate results (some of the sites are marked as target), and the movement from site to site occurs through the use of transformations from a given set. In M.L. the solution to the problem is finding a path from the initial platform of the maze to one of the target sites. In this case, the labyrinth is considered fully specified. (See also relational model.)
1. A model related to the fixation of certain knowledge about natural language.
2. Description of the object in terms of linguistic variables and arguments about them.
A model based on the expansion of a formal system, within which procedures are introduced to change all or part of the elements of a formal system depending on the tasks to be solved. M.L.L. Often used as a way to specify a Kripke model.
Model of knowledge representation, which is based on a formal system.
The way of displaying in the memory of the intellectual system of knowledge about the environment (See also Knowledge Model, Conceptual Diagram.)
The model underlying the process of learning a person or a technical device. There are two types of M.O. - descriptive and regulatory. Descriptive M.O. is extracted from the description of the process of the activity by which the person or system should be trained. This extraction can occur in a variety of ways. The most famous of them is based on the learning procedure with examples. Normative M.O. set in advance. Often, learning of the normative type is called learning with a teacher.
Description of the body of knowledge about how communication is organized between the user and the intellectual system. Usually in M.O. includes a user model and a dialogue flow model. If communication occurs in a professional natural language, then the results obtained in the theory of speech acts are used to build a user model. In other cases, the procedures for the exchange of graphic information through the display screen can be applied.
A model in which new facts and patterns can be added to the process of functioning of an intellectual system
A model (technical or software) that reproduces certain types of object behavior under certain environmental conditions (overcoming obstacles, reacting to external influences, making decisions, etc.). Mn It is used both in the study of the real behavior of biological systems and humans, and in the development of intelligent robots (planning their automatic behavior).
The set of knowledge about the features of the user's work with the system, its intentions, goals and requirements, which is stored in the memory of an intelligent system. Mn helps the system to organize an effective dialogue with the user, creates psychological comfort for him.
A data description model in which all relationships are specified by rows of tables whose columns are labeled with attribute names. Tabular presentation of data is often convenient. This ensured the widespread dissemination of relational databases. At the heart of M.R. Special predicate calculus.
The knowledge representation model based on the semantic network.
Classification model that allows identifying current situations as known systems. M.S. used, for example, in situational management.
In artificial intelligence, a set of procedures and declarative descriptions, with the help of which the intellectual systems imitate the part of human conscious activity that can be verbalized. In psychology, the term "consciousness" is interpreted more broadly. It includes, for example, the ability of the subject with consciousness to self-observation (self-awareness), reflection and activity.
The model of behavior based on the principle of the black box. In M.S.P. It considers a finite set of stimuli that can be perceived by the subject or fed to the output of an artificial system, and the rules for correlating these stimuli with the reactions issued by the subject or system. Internal processes linking stimuli and reactions are not analyzed and not taken into account. M.S.R. It is used in intellectual systems at the level of reproducing the simplest behavioral reactions to stimuli coming from the external environment.
Part of the communication model, which is a description of the types and structures of the dialogue that are in the possession of an intellectual system and which the user can use in communicating with this intellectual system. M.T.D. is given in the form of either hard rules, or automaton grammar, or a script. (See also User Model.)
A formal description in some logical language of the structure of an object. (See also formal system.)
In linguistics - a formalized representation of knowledge about the language. As a rule, it includes morphological, syntactic, semantic and pragmatic components, which can also be divided into more fractional components.
Inference rules, in logic which states: "If A and A®B are derived, then B is derivable."
The rule of inference in logic, which states: "If A and BA is true, then B is true." This rule is used in the resolution method.
The property that is typical for inference in a closed formal system and in a closed knowledge base and that the previously derived statements do not lose their truth when expanding the set of premises for output.
The property used in databases and knowledge, which consists in the fact that if two information units are interconnected by relations of the type "genus-species" or "class-element", then the information common to all species included in the genus, or for all elements included in the class is contained in the information unit of a higher level and, if necessary, is inherited by a unit of a lower level. N. allows you to eliminate duplication in storing information in databases and knowledge.
Direction in research on artificial intelligence which is characterized by the use of reproducing in intellectual systems of processors, inherent in biological objects, structures and functions similar to the structures and functions of these objects. Within this framework, formal models of neurons were created, on the basis of which networks are built, allowing to solve problems of pattern recognition, classification, stimulus-reactive behavior, etc. The complication of the structures of formal neurons leads to structures with wide functionality. They are often called neurocomputers. Examples of a neurobionic device are the perceptron.
See computer neurobic.
The element whose operation is described by the function
/
| 0 when SAi X i - S B i Z i y = <
| 1 when S A i X i - S B i Z i > = h
\
Here y is the
продолжение следует...
Часть 1 Dictionary of Artificial Intelligence
Часть 2 DENOTAT - Dictionary of Artificial Intelligence
Часть 3 Nemonotonnost with the withdrawal - Dictionary of Artificial Intelligence
Часть 4 SYNTAX - Dictionary of Artificial Intelligence
Comments
To leave a comment
Artificial Intelligence. Basics and history. Goals.
Terms: Artificial Intelligence. Basics and history. Goals.