Lecture
Special terminology on artificial intelligence and intellectual systems began to take shape in the 60s of the twentieth century. The first stage in the formation of terminology is always characterized by the presence of many synonymous terms that are used by various schools and groups of specialists. At this stage, the terms quickly arise and some of them also quickly disappear. By the mid-70s, terminology in the field of artificial intelligence began to be established. There are terms that recognized the vast majority of specialists. All of these terms (with rare exceptions) are English-speaking in origin, since it was in the United States that intensive research was conducted in this area. Finally, the main terminology entrenched in the first half of the 80s. From this period dictionaries and encyclopedic reference books on artificial intelligence began to be published. The difficulty of creating an explanatory dictionary on artificial intelligence is related to the interdisciplinary nature of research in this area. Since artificial intelligence uses methods traditionally developed in logic, psychology, linguistics, cybernetics, discrete mathematics and programming, there are quite a few terms and other sciences. The discussion of the explanatory dictionary showed that there are two extreme cases: include terms that are used only in artificial intelligence and are not borrowed from other sciences, or include all terms related to artificial intelligence, if they are found in the works of specialists in this field. . The compilers of the dictionary made an interim decision, which, of course, immediately set before them the problem of boundaries. Using the collective experience of specialists from a number of countries and the USSR, the compilers attempted to define this natural boundary by associating it with the frequency of occurrence of terms borrowed in artificial intelligence.
abduction abduction
abstraction abstraction
data abstraction data abstraction
automaton machine
endless machine
probabilistic automaton
deterministic automaton
initial automatic
cellular cellular automaton
final automatic machine finite automaton
automatic linear-bounded automaton
automatic push-down automaton
nondeterministic automaton
variable structure machine
sequential automaton
stack machine
automatic stochastic stochastic automaton
office automation office automation
automated control system (ACS)
ACS enterprise (ASUP)
Process control system (PCS)
axiom axiom
speech act act
actant
actor actor
algorithm algorithm
wave algorithm
genetic genetic algorithm
analysis
cluster cluster analysis analisys
morphological analysis morphologic analisys
protocol analysis protocol analysis
syntactic analysis syntactic analysis
scene analysis scenary analysis
parser syntactic parser
analogy analogy
anaphora anaphora
argumentation argumentation
computing system architecture
computer architecture computer architecture
pipeline architecture
architecture parallel architecture
architecture stream data flow architecture
association association
atom atom
attribute attribute
data base database
database hierarchical hierarchical data base
relational relational data base
network database
extensional data base extension database
knowledge base knowledge base
closed knowledge base closed knowledge base
knowledge base intensional intensional knowledge base
open knowledge base open knowledge base
backtracking backtracking
Socratic conversation Socratic dialogue
behaviorism behaviourism
validation validation
verification verification
video processor
perception perception
visual perception
perception of tactile perception information
training selection teaching selection
inference output
output abductive abductive inference
output probabilistic probabilitic inference
output natural common-sense inference
inductive inference output
intuitionistic intuitionistic inference
linear output linear inference
logical output logical inference
knowledge output based on knoledge-based inference
output non-monotonic non-monotonic inference
fuzzy inference fuzzy inference
reverse output
analog output
output plausible plausible inference
output direct direct inference
call by pattern call by pattern
saying proposition
atomic proposition
text generation text generation
hermeneutics hermeneutics
hyper-event hyper-event
hypothesis hypothesis
compactness hypothesis
deep structure (sentences)
grammar
automaton grammar
grammar context sensitive
context free grammar
context-related grammar
grammar matrix grammar
non-shortening grammar
case grammar case grammar
grammar network network grammar
formal grammar
graph graph
graphics dynamic animation
cognitive graphics cognitive graphics
graphics machine
action action
decomposition of tasks problem decomposition
denotatum denotatum
tree inference tree
binary tree
dependency tree
decision tree
tree components consistency tree
goal tree
designat designig
descriptor handle
default default
clause
side clause
empty clause clause
Horn clause Horn clause
disjunction disjunction
discourse discourse
dissonance cognitive cognitive dissonance
constructive proof
proof of theorem theorem proving
domain domain
blackboard message board
law of the excluded middle
the law of removing double negation
knowledge knowledge
declarative knowledge
knowledge of the subject area
pragmatic knowledge pragmatic knowledge
knowledge of procedural knowledge
knowledge heuristic heuristic knowledge
expert knowledge
attribute value attribute
default value
vision computer vision
AND / OR graph AND / OR graph
identification identification
knowledge identification
knowledge acquisition knowledge acquisition
AI programming AI programming
illocutionary potential
illocution illocution
imitation of intellectual behavior
imitation of thinking processes
implication implication
induction
induction incomplete (empirical) empirical induction
induction complete (mathematical) complete induction (mathematical)
knowledge engineer
knowledge engineering knowledge engineering
artificial intelligence artificial intelligence
interview interview
interpretation of interpretation
interface
interface natural language natural language interface
intelligent interface
artificial brain
knowledge source
calculus calculus
propositional calculus
genten calculus
calculus logical logical calculus
predicate calculus predicate calculus
first order predicate calculus
calculus propositional propositional calculus
calculus situational calculus
cognitive map cognitive map
causation causation
quantifier quantifier
quantification quantification
universal quantifier
existential quantifier
classification
clustering clusterization
clause
cogitology cogitology
cognitive science cognitive science
component declarative component
concatenation concatenation
Kelly Kelly construct construct
concept concept
conjunction conjunction
linguistics computational linguistics
computer linguistics computer linguistics
lips lips
literal
logic logic
probabilitic logic probabilistic logic
belief logic belief logic
logic temporal logic
second order logic second order logic
binary logic
action logic operational logic
deontic logic deontic logic
dynamic dynamic logic
common sense logic common-sense logic
inductive logic
intuitionistic logic intuitionistic logic
causal logic causal logic
command logic command logic
constructive logic
mathematical logic mathematical logic
multi-valued logic multi-valued logic
monotonic logic monotonic logic
non-monotonic logic non-monotonic logic
odd logic fuzzy logic
logic of norms
deontic logic evaluation logic
first order logic first order logic
logic propositional propositional logic
logic spatial spatial logic
pseudophysical logic pseudophysical logic
blurry logic fuzzy logic
default reasoning logic
logic epistemological epistemiological logic
lok
locution locus
lambda calculus lambda-calculus
machine abstract abstract machine
database machine data base machine
base machine knowledge base machine
virtual machine virtual machine
parallel output machine parallel inference machine
machine Post machine
connection machine
Turing machine Turing machine
data flow machine
machine intelligence
menu menu
plausibility measure
meta-knowledge meta-knowledge
meta-production meta-production
metaphor metaphor
meta language meta-language
branch and bound method branch-and-bound method
interview method interview method
back wave method
direct wave method
inference mechanism
inheritance mechanism
MIMD architecture MIMD architecture
a lot of fuzzy fuzzy set
model model
associative model associative model
computational model computational model
closed closed model
knowledge model knowledge model
cognitive model cognitive model
model conceptual conceptual model
Kripke model Kripke model
labyrinth model labyrinth model
linguistic model linguistic model
logical-linguistic model logical-linquistic model
logical model logical model
world model
learning model
communication model
open model open model
behavior model
user model
relational model
network model
situational model
model of consciousness
SR-model stimulus-reaction model (stimulus-reaction model)
dialogue flow pattern
formal formal model
language model
modus ponens modus ponens
modus tollens modus tollens
output monotony
inheritance hireditance
neurobionics neurobionics
neurocomputer
neuron formal
non-monotonous at output
uncertainty uncertainty
linguistic uncertainty
incompleteness incompleteness
undecidability algorithmic algorithmic nonresolvability
new information technology
subject area
subject area poorly structured ill-structered subject area
well-structured subject area well-structured subject area
problem area problem area
generalization of knowledge
inductive generalization
shell
justification argument
natural language processing
image processing image processing
parallel processing parallel processing
signal processing
image image
pattern pattern
learning learning
example learning learning from examples
communication communication
combination of evidence combination of evidences
explanation explanation
integrity constraint
revival
justification excuse
knowledge base debugging
semantic debugging
syntax debugging
relation relation
anti-reflexive antireflexive relation
relation antisymmetric antisymmetric relation
antitransitive relation
virtual virtual relation
temporal relation
action relation
attitude intensional relation
causal relation
fuzzy modeling relation
non-reflexive non-reflexive relation
non-symmetric non-symmetric relation
non-transitive relation
spatial spatial relation
relevance relation
reflexive relation
semantic relational relation
symmetric relation symmetric relation
tolerance relationship
transitive relation
functional relationship functional relation
equivalence relation equivalence relation
relational extensional relational relation
negation negation
negative logical logical negation
deep case
case Fillmore Fillmor case
associative memory associative memory
virtual virtual memory
iconic iconic memory
papline architecture pipeline architecture
translation machine machine translation
linguistic variable linguistic variable
propositional variable propositional variable
bound variable variable
perlocution perlocution
perceptron perceptron
perception perception
pixel pixel
pictogram
planning planning
planning activity planning
hierarchical planning
distributed planning
strategic planning strategic planning
planning tactical tactic planning
planner planner
Bayesian approach Bayesian approach
search search
search associative search
depth-first search
search in task space search in problem space
search in the state space search in state space
wide search breadth-first search
search ascending
search information information search
search downward
pattern pattern matching
best-first search best-first search
search random
search first deep
first search for breadth
natural language understanding
concept concept
hypothesis generation automated automated hypothesis generation
text generation text generation
data flow
de morgan de morgan rules
rule rule
inference rule output rule
compositional compositional inference rule
syntactic rule
predicate predicate
data representation
knowledge representation knowledge representation
extensional representation representation
presupposition presupposition
resolution principle
knowledge acquisition acquisition
game game program
heuristic heuristic program
programming programming
programming logical logical programming
object-oriented object-object programming
functional programming
programming heuristic heuristic programming
production production
proposition
task space
space Osgood Osgood space
semantic space semantic space
state space state space
target space
inconsistency absolute absolute inconsistency
inconsistency model model inconsistency
frame prototype frame
refutation procedure
attached procedure
asynchronous process asynchronous process
associative processor associative processor
data base processor
linguistic processor
logical logical processor
logic processor inference processor
matrix matrix processor
character processor simbolic processor
psychology cognitive cognitive psychology
solvability algorithmic resolvability
pattern-recognition pattern recognition
speech recognition speech recognition
semantic distance
reasoning reasoning
reasoning autoepistemic autoepistemic reasoning
reasoning hermeneutic hermeneutics reasoning
common-sense reasoning reasoning
nonmonotonic reasoning non-monotonic reasoning
reasoning by analogy reasoning by analogy
reasoning by association
default reasoning default reasoning
reasoning plausible reasoning
resolvent resolvent
resolution resolution
problem solver
Kelly repertoire lattice repertoire lattice of Kelly
RISK-architecture RISC-architecture
autonomous robot autonomous robot
robot integral integral robot
intelligent robot
garbage collection
evidence of evidence
sequent sequence
semantics semantics
semantics situational situational semantics
semiotics semiotics
network
associative network associative network
output network inference network
causal network
transition network extended argued transition network
Petri Petri network
causal network causal network
semantic network
network semantic intensional intensional semantic network
network semantic extensional extensional semantic network
network connecting connectional network
syllogism syllogism
SIMD-architecture SIMD-architecture
syntax syntax
program synthesis automated program synthesis
synthesis programs deductive inductive program synthesis
program synthesis inductive program synthesis
transformational program synthesis
text synthesis
computer-aided design computer-aided design
axiomatic axiomatic system
question-answer question-answering system
deductive system
belief system belief system
natural language system natural language system
inductive system
intelligent system
intelligent teaching system
intelligent learning system
interactive system
quasiaximatic system
multiprocessor multiprocessor system
learning system learning system
explanatory system explanation system
knowledge-based system
rule-based system
knowledge representation system
production system
production system production system
fifth generation computing system fifth generation computer system
control system automated computer-aided control system
database management system
knowledge base management system
production management system
process control system
formal formal system
frame system
expert system expert system
expert expert tool system
consequence logical logical consequence
slot slot
event
pattern-matching pattern matching
knowledge component intensionalnaya
extension component of knowledge
associative associative list
knowledge engineering tools knowledge engineering tools
anaphoric reference anaphoric reference
inference control strategy
deep structure
cognitive cognitive structure
homogeneous structure homogeneous structure
case structure
cognitive cognitive structure
judgment judgment
entity entity
conceptual conceptual scheme
script script
creativity computer art
Church thesis Church thesis
axiomatic theory
logical theory
speech theory
term term
data type
data type abstract abstract data type
universum universum
Herbran Universum universbra
unifier unifier
most general unifier
unification unification
situational control
facet
focus focus
form prefix normal prefix normal form
atomic formula atomic formula
closed formula closed formula
valid valid formula
opening formula
fractal fractal
frame frame
frame prototype frame
case frame
prototype frame prototype frame
frame instance frame example
output function
belief function belief function
transition function membership function
membership function membership function
Skolem function Skolem function
inference chain inference chain
absolute scale absolute scale
metric scale
Osgood scale Osgood scale
relative relative scale
blurred fuzzy scale
scale topological topological scale
universal universal scale
Neurobiotic computer neurobionical computer
heuristics heuristics
automatic tongue
query language query language
context free language
context related language
knowledge representation language
language of knowledge representation
production language production language
sequence language
frame language
gray box
box black black box
A plausible conclusion from private to private.
The process of exclusion of a single, random or irrelevant work for the subsequent steps. A. always takes place in the presentation of data and knowledge about the external world of intelligent systems. A. Used in the synthesis of knowledge, conducting reasoning and planning expedient activities. A. is a means of forming concepts.
1. Use in converting variables only the operations connecting them without taking into account the internal representation of variables.
2. A programming methodology in which a program is described as a collection of abstract data types. (See also Abstraction)
An abstract machine that transforms a sequence of input characters into a sequence of output characters. Depending on the number of internal states of the memory A. the finite A. and the infinite A .; depending on the unambiguity or ambiguity of the formation of the output sequences - deterministic A. and non-deterministic A .; depending on the characteristics of the structure, store A., stack A., cellular A.
An automaton whose set of internal states is countable, in particular, the Post machine and the Turing machine.
A special case of a stochastic automaton, when the structure of the automaton remains unchanged for any results of its operation.
An automaton whose set of input symbols and the internal state uniquely determine the set of output symbols and the internal state of A.D. in the subsequent tact of work.
Automatic machine with a pre-fixed internal state at the beginning of work.
A homogeneous structure consisting of cells, each of which contains a finite state machine, in the general case A.K. It has four entrances from neighboring cells and four exits leading to them. All machines in the cells are the same. AK allows you to simulate parallel asynchronous processes. In particular, using AK, one can model the self-organization of various space-time configurations.
An automaton whose operation is determined by two functions:
y (t + 1) = F 1 (x (t), y (t)),
z (t) = F 2 (x (t), y (t)).
The first function sets the state change of the automaton to discrete clock cycles of time t and is called the transition function; the second is the output signals of the machine and is called the exit function; x, y, and z are sets of binary vectors of fixed length, i.e. finite sets. The mathematical model of AK can be an automaton grammar with the help of which an automaton language is generated.
Private view of the Turing machine, which has a finite length at each time instant. If it is necessary to shift the control head beyond the edge of the tape, the tape is built up to the final segment that the head needs. Linearly-bounded automata correspond to context-dependent grammars, generating context-dependent languages.
This is a special case of a stack automaton, which can read only the information that was written to the stack last.
An automaton with a set of input symbols and an internal state specifying an alternative choice of a set of output symbols and / or an internal state A.N. in the subsequent tact of work. A special case of A.N. are probabilistic automaton and stochastic automaton.
See Automaton Stochastic.
A finite state machine described in a language of sequences that defines automaton functions. Each such system can be associated with a typical AS structure. consisting of a register (interconnected flip-flops), a matching circuit and two diode arrays, one of which serves to implement the transition functions of the automaton, and the other functions of the outputs.
An automaton whose memory is organized in the form of a stack, in which a sequence of input symbols is stored with preservation of the order of their arrival. Reading information from the stack is made by the position number in the stack. A special case of A.S. is a vending machine. A.S. used when generating context-sensitive languages with a given depth of contexts, which leads to its use in linguistic processors.
An automaton, in which, instead of the transition and output functions, in the general case, probability distributions of a discrete type are specified. For transitions, the probabilities H ij are given, which characterize the probability of a change of state with number i to the state with number j, and for the output, the probabilities Q ij , which characterize the appearance of output with number j, if the current state of the automaton is i. AC is often used to describe the process of adaptation to the environment in which it operates. Depending on the success or failure of AC actions, H ij and Q ij are recalculated, which leads to AC adaptation if the medium is stationary.
A set of tools that automates the process of office work at the level of one employee. HELL. allows you to store a set of documents in the computer's memory, scroll through document folders on the display screen, correct documents, print and put new documents in folders, destroy unnecessary ones, etc. Using computer graphics allows you to display on the screen the usual type of documents.
A set of tools for automated management of organizational and organizational-technical systems, including a set of subsystems that implement all the functions necessary for planning, operational management and reporting. When ACS is being intellectualized, it may include expert systems, intelligent information systems, databases and knowledge, and a natural language interface.
Complex software and hardware for automated enterprise management. The main tasks of the automated control system are: calendar and long-term planning, organizational and operational management, sales and supply, financial accounting, statistical reporting. The automated control system includes a system for collecting and processing information, as well as such system intellectualization tools as databases, knowledge bases, expert systems and a natural language interface. The lower levels of automated control systems (its parts) are automated process control systems.
A complex of software and hardware that directly controls the process (production) process. The main tasks of the process control system are tracking and adjustment of the technological process, the solution of operational dispatch tasks and the formation of a higher level of control (ASUP). В состав АСУТП могут входить такие средства интеллектуалиации системы, как системы визуализации информации, экспертные системы и естественно-языковый интерфейс.
Утверждение, которое априорно считается истинным.
Произнесение говорящим некоторого сообщения при наличии слушающего и конкретного ситуативного окружения. Изучение Р.А. позволило сформулировать требования к успешности коммуникации при общении людей друг с другом и людей с искусственными системами. Изучается в теории речевых актов.
Термин А. связан с определенной синтаксической концепцией, согласно которой элементарные высказывания расчленяются на функции (субъект, объект, предикат) и предикат рассматривается как ядро высказывания.
Специальным образом организованная процедура, характерная для объектно-ориентированного стиля программирования. Особенность А. состоит в том, что он самостоятельно включает в работу (активизируется), когда выполняются условия активизации. После окончания работы А. передает полученные им результаты другим А. Использование А. обеспечивает децентрализованное и максимальное параллельное решение задач.
Система предписаний, задающая последовательность действий, с помощью которых за конечное число шагов находится решение всех задач определенного класса или выдается сообщение об отсутствии решения.
Organization of the computational process on a network structure, for example, a semantic network. It is assumed that at the vertices of the network there are operators that need to be processed, and the arcs (edges) of the network describe possible connections between operators. A wave of processing propagates from each operator along all arcs emanating from it, taking into account the requirements imposed on the organization of wave propagation A.V. A.V. used when there is a set of devices capable of parallel and asynchronous processing of the network structure. A.V. showed high efficiency in solving problems of graph theory and networks, as well as in search and sorting procedures (for example, when searching on a model in knowledge bases).
The organization of a process resembling evolution in nature. Alternative solutions plyustiruet population. Mechanisms similar to mutation and crossing-over are used to select a solution. Alternative solutions are compared with each other and the “surviving” solution that acquires the maximum weight in the “population”, A.G. used in learning models.
A method, a type of research in which a real or conceivable object is dismembered into its component parts (elements) and these elements and the connections between them are studied.
The partitioning of a set of objects into clusters (compact groups of objects) in a specially designed space, whose metric is such that objects close to each other from the point of view of this metric fall into one cluster. Clusters can be distinguished in many ways by the theory of A.K.
Text processing, giving information about the morphological characteristics of each word of the text.
Conducting a protocol with an expert in acquiring knowledge, in which the entire dialogue is recorded in some way (magnetic recording, transcribing, etc.).
1. Checking an expression for its occurrence in a set of constructed expressions.
2. When processing natural language under A.S. the construction of the syntactic structure of a sentence in some natural language is understood.
In intelligent systems that implement the functions of understanding texts in a limited natural language, A.S. carried out in two stages. The first one builds a superficial syntactic structure in which the parts of speech and the relationship between them are involved. At the second stage, a transition takes place to the deep syntactic structure, which is essentially a cognitive structure associated with the reflection of reality in the text in a natural language. For A.S. Created software systems - parsers.
A complex of models and methods that allows to display three-dimensional visual scenes in the system’s memory in intelligent systems (for example, in intelligent robots). When projecting scenes into memory, a transition to their formal description takes place. At the same time, it is necessary to correctly recognize the objects participating in the scene, determine their location in depth, supplement the enclosed areas of objects, etc.
Means for checking the correspondence of input chains of a given language to its grammar. Classification AS, as a rule, is based on the classification of the corresponding formal grammars.
The relationship between two objects, processes, events, or situations at the level of the similarity-difference relationship in the knowledge base.
Repetition of the same word or phrase within the same sentence or adjacent sentences in the text. Often, the repetition does not use the word or phrase itself, but the so-called anaphoric word (most often a pronoun). For example, in the text "Peter went to school. She was on the edge of the village" "she" is an anaphoric word for the word "school". Such anaphoric references make it difficult to automatically analyze texts for understanding them in intelligent systems.
The process of proving the truth of a statement with the involvement of facts from which the truth of a given statement follows or which increases confidence in its truth A. is close to substantiation.
See Computer Architecture.
A collection of generalized information about the structure of the main blocks of a computer, their functioning and information and control links between them.
The architecture of a computing system consisting of a chain of parallel processors that interact in such a way that the output of one of the processors of the chain is connected to the input of the next processor. In this case, the processor consists of consecutive elements, each of which implements partial processing of a command (command selection, operation code decryption, address arithmetic, selection of operands, operation execution) and the next comma *** begins to execute earlier than the previous one. A.K. It has high performance in the implementation of algorithms characterized by the use of identical sequences of operators to different source data (in the so-called mass data processing systems).
The architecture of the computing system in which the same or different operations are performed simultaneously on several data groups. See also SIMD architecture and MIMD architecture.
The computing system architecture is oriented towards controlling the computational process using a data stream or request stream. In the first case, the execution of each operation is initiated by the contents of its operands: the sequence of command execution is not specified in advance. Commands as operands are not addresses of memory cells, but instructions whose results are operands of this command. Such an organization of computers is focused on the use of functional programming (LISP and the like).
A link between two information units in the knowledge base, established on the basis of some measure of proximity, defined on the set of information units stored in this base.
See atomic formula.
A unique name assigned to the value domain of an information item.
The set of software tools that provide search, storage and recording of information units of a given structure (data) in the computer memory.
A database in which relations between the information units are entered as "Element-class", "type-subtype", etc., which are used to form hierarchical classifications of information units stored in the database.
The database, in which information units are interconnected by one-to-one relationships, has attributes and uses a table-like entry to represent the relationships.
A database in which information units are interconnected by one-to-one, one-to-many, and many-to-many relationships.
A database that stores only constant facts about the outside world.
The set of software tools that provide search, storage, transformation and recording in computer memory of complexly structured information units (knowledge).
The knowledge base, the contents of which in the process of operation does not change. The logical conclusion in such a base is equivalent to the conclusion in the formal system and has the property of monotony, i.e. previously derived statements remain true for the entire period of operation of B.ZZ.
The knowledge base, which describes the general patterns characteristic of a certain problem area, as well as ways of setting and solving problems in this area.
The knowledge base allows in the process of its operation to replenish the contents of the base and remove knowledge from the base. The property of openness leads to the fact that the output in such a base is non-monotonic, i.e. the truth of the statements derived in it may change in the process of the system with such a base.
Return procedure when searching on some structure (for example, search in a decision tree or in a maze). When moving along the structure, it is often necessary when the chosen path was unsuccessful or a dead end in returning to the place of branching of the search process. To speed up the possibility of returning to the last branch point, its coordinates should be stored in memory. To store the aggregate nested in each other by the precedence of branch points, special stack registers are used.
A specially organized dialogue, during which one of the participants proposes a thesis, and the second consistently puts forward objections to the thesis, to which the participant who proposed the thesis answers only: "I agree" or "I do not agree." Purpose B.S. is learning the techniques of logical reasoning.
Reducing the creation of purposeful behavior associated with a pair of "stimulus-response". B. is characterized by considering the subject as a black box. A number of models of artificial intelligence was built on the basis of the approach declared in B.
Evaluation of a software product in terms of compliance with all requirements for it.
Evaluation of the correctness of the source data for the production of all requirements for the future product, and for its production.
Specialized processor designed for the effective implementation of image processing algorithms. A feature of the interaction of the raster display with the special processor is the correspondence of each point of the image of one or several memory bits in the address space of the special processor. By changing the contents of the memory cells, the program changes the image on the screen.
Reflection of the surrounding situation and its elements in the interaction of the human sense organs or receptors of the artificial system with the external environment. W. provides directly-sensual orientation in the environment and generates a stream of input information for subsequent processing by a person or an artificial system. For intelligent systems, the most important types of perception are the perception of visual information, the perception of tactile information and acoustic information (speech recognition).
Processing of signals entering the intelligent system (intelligent robot) from sensors of visual scenes. As sensors for flat images, photodiode arrays are often used, but photographic equipment is also used. For three-dimensional scenes, a television camera is most often used as a sensor. When V.Z.I. the image is cleared of noise and distortion, a flat image analysis or scene analysis for a three-dimensional image, image transcoding and transfer to the knowledge base or solver.
In intelligent robots, the processing of signals from sensors measuring the kinematic characteristics of robot effectors, and from special sensors measuring forces associated with taking objects or resting on the ground. After pre-processing, the information obtained enters the activity planning system and is used to develop control actions on the environment or on the robot.
A set of examples and counterexamples to form decision rules. Included in the training table.
Getting new information units from previously known. A special case is the logical conclusion.
Conclusion based on abduction.
The conclusion that every expression used in it has a likelihood estimate in the form of the probability that it is true. When V.V. special procedures are used to calculate the probability of the true value of the resulting expression for the probabilities of the premises used in the derivation.
The conclusion obtained on the basis of "common sense". V.E. It can either correspond to a logical conclusion in some formal system (but be obvious to a person), or rely on considerations that do not fit into the strict framework of the formal system.
Conclusion "from private to general." Allows, on the basis of generalization of particular examples of a certain phenomenon, to put forward a hypothesis about the existence of a general pattern. In intelligent systems that use VI, a mechanism works that makes it possible to attribute a likelihood estimate to it when forming a hypothesis (for example, the probability that this hypothesis is true). IN AND. is a means of obtaining new knowledge in intelligent systems.
The conclusion characteristic of intuitionistic logic, which does not use, in particular, the law of the removal of double negation and the law of the excluded middle.
The sequence of clauses in which the initial clause belongs to the original set, in each intermediate clause is the resolvent of the preceding in the already constructed part of the sequence of clauses and some lateral clause.
1. A sequence of reasoning, leading from premises to a corollary using axioms and inference rules.
2. Output result.
A conclusion that uses expressions stored in the knowledge base of V.N.Z. can be reliable if these expressions are reliable, or plausible, or provided with likelihood estimates. As a rule, the procedures V.N.Z. include the search for the necessary knowledge for output, i.e. sample search procedure.
Conclusion in which the property of monotony is violated in the output.
A conclusion that uses fuzzy quantifiers or membership function values. For fuzzy quantifiers, the derivation rules determine the quantifier that should be assigned to the result for given values of the quantifiers of the premises. When using the values of membership functions, the output rule determines the value of this function for the result based on the values of the membership functions of the parcels.
The conclusion that the search for evidence begins with a target statement. The conditions under which the target statement is deducible are determined. These conditions are accepted as new targeted statements and the search process continues. IN. it ends when all the next conditions turn out to be axioms or the process of conditions terminates without leading to axioms. IN. widely used in intelligent systems when searching for solutions.
Conclusion based on the transfer of reasoning from the study area to another area similar to the study. If there is a pin A ® B, and the region in which A is defined is homomorphic to the region where C is defined, and the region where B is defined is homomorphic to the region where D is defined, then the pin A ® B generates the pin C ® D. V.P. .BUT. There is a special case of a plausible conclusion.
The conclusion at which each step is accompanied by the calculation of the assessment of the reliability of the assertion obtained. Special cases of V.P. is, for example, probabilistic inference and inductive inference.
Conclusion leading from the source axioms to the target expression. When V.P. because of the ambiguity of the choice of the applicable axioms and the derivation rules, a decision tree is formed and the process of finding a chain leading from the initial axioms to the target expression is overdraft. The standard procedure used when traversing a decision tree is the return procedure - backtracking.
A way to search for information in databases, knowledge bases or in computer memory. In contrast to the search by address storage, VPO. assumes an associative search by content search query. (See Search by sample).
A logical expression, in relation to which it can always be argued that it is either true or false.
The statement, the structure of which is not dissected further. In formal systems V.A. match the basic elements. Interpretation of the truth of V.A. is set for the formal system from the outside and on this basis the interpretation of all correctly constructed formulas of the formal system is determined.
The process of generating text, including the selection of a fragment of the internal representation that will be included in the text; formation of a discourse scheme, i.e. the sequence in which the information should be presented; filling the discourse scheme with language expressions. When G.T. It also takes into account the focus of attention and the prevention of communication errors. (See Text Generation. Text Synthesis).
The semantics section that studies ways of detecting the content of a text is explicitly expressed in it. Search for hidden content in the text occurs as a result of reference to knowledge that is relevant to this text.
A specially organized description of a typical situation (theft, fights, shopping, etc.). G. can be represented in knowledge bases in various ways, for example, in the form of scenarios.
Частично обоснованная закономерность знаний, или для связи между различными эмпирическими фактами, или для объяснения факта или группы фактов. В интеллектуальных системах Г. порождаются в процессе обучения систем (в частности, при обучении на примерах).
Предположение о том, что образы в пространстве признаков группируются из изображений (точек пространства), которые могут быть отделены друг от друга гиперповерхностями простого вида. Гипотеза Компактности используется при распознавании образов, когда применяется принцип разделения.
См. Структура глубинная.
Совокупность правил формирования правильных предложений в рамках рассматриваемого языка.
Формальная грамматика, у которой правила вывода имеют вид
b 1 A 1 ® b 2 A 2 , b 1 A 1 ® b 2
где A 1 , A 2 - нетерминальные символы; b 1 , b 2 - терминальные символы. Каждая Г.А. порождается некоторым конечным автоматом. И каждый конечный автомат задает некоторую Г.А.
См. Грамматика контекстно-связанная.
Формальная грамматика, для которой существуют такие цепочки что и имеют место правила и К.С. обладают свойством сохранения длины цепочки. Цепочки, получаемые после применения любого правила, либо сохраняют длину исходной цепочки, либо увеличивают ее. Г.К.С. порождаются линейно-ограниченными автоматами. И для каждой Г.К.С. может быть построен воспроизводящий ее линейно-ограниченный автомат.
A formal grammar, in which the rules of inference are fixed in the order of application of the set of rules of inference of ordinary grammars. These sets, called matrix inference rules, can overlap with each other according to the general inference rules included in the matrix inference rules. Gm used in the description of parallel processes and programming languages for devices that provide parallel execution of programs.
See Grammar context-related.
Grammar of the predicate-argument structure of the sentence. Arguments of such a structure are names for which you can specify deep cases (generalized relations between the content of the verb and the content of one or another of the nominal groups).
The grammar of extended transition networks is a subclass of transformational grammars. Currently under GS understand the ordered triple F s = , where V is the description of the lexicons of the language being processed; Z - description of non-standard functions that increase processing efficiency; N is the description of the extended transition network (a special graph with which the language analyzer is presented).
The quadruple , in which S is the axiom of GF; A is the set of non-terminal symbols; B is a set of terminal symbols; P - rules of inference. The objects with which GF works are chains consisting of terminal and non-terminal symbols. Inference rules have the form, where the string of characters. It contains at least one non-terminal character. Functioning gf always starts with a string consisting of a single symbol S. Applying the rule to a string is to replace all occurrences (or only the leftmost entry) in on. The process ends when none of the inference rules apply to this chain. Such final chains enter the language generated by a given GF. In syntactically correct gf the language includes only those chains that consist entirely of terminal characters. G.F. widely used in syntactic models for natural languages and linguistic processors. They are a special case of formal systems. Depending on the restrictions imposed on the structure of the rules of inference, various types of GF are distinguished.
The pair (X, R), where X is a set whose elements are renamed and are called vertices; R is a binary relation defined on X. If there is a relation R between vertices x 1 О X and x 2 О X, then the triple x 1 R x 2 is called an edge G. If the relation R is asymmetric, then x 1 R x 2 is called an arc T A G. with edges is called undirected, and with arcs it is oriented. G. are widely used in artificial intelligence models.
Direction in computer graphics, which develops techniques and procedures for reproducing moving scenes on the display screen. In gd many tools developed in animated (animated) cinema, as well as a number of mathematical techniques are used.
Direction in computer graphics, which links the views that appear on the display screen, with cognitive processes occurring in solving problems. G.K. allows you to visualize the decision process. With a well-thought-out visualization system, images arising in dynamics on the screen can help the user solving the problem in interactive mode, to see those patterns or solutions to the problem that were previously not available to him. With the development of G.K. they pin great hopes on improving the efficiency of solving problems, since the user's thinking can significantly speed up the process of finding a solution and give rise to new ways of finding it. G.K. requires special representations in the knowledge base of the corresponding images on the display screen, and procedures for correlating these representations with traditional cognitive structures.
1. A set of software tools for issuing to the display images in graphic form of intermediate and final results of solving problems and for working with graphic images.
2. Direction involved in the development of these tools.
Unit of the activity process, activity aimed at achieving a specific goal. D. may be internal, aimed at converting information of the internal intelligence system, or external, directed to the external environment (message to the user, movement of the autonomous robot arm, etc.).
Partitioning the task into subtasks with the subsequent partitioning of these subtasks to obtain basic (elementary) tasks for which the solution is known in advance. D.Z. It is used in the intellectual systems when creating automatic programming systems and when planning behavior in the task space. In a more general sense, D.Z. can serve to understand the dimension of the problem being
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.