semantic field. Recordsets of links between them
Semantic field - a set of linguistic units united by some common (integral) semantic feature; in other words, having some common nontrivial value component. Initially, the role of such lexical units was considered as units of the lexical level - words; later, descriptions of semantic fields appeared in linguistic works, including also phrases and sentences.
One of the classic examples semantic field can serve as a color naming field, consisting of several color rows ( red– pink – pinkish – crimson; blue – blue – bluish –turquoise etc.): the common semantic component here is "color".
The semantic field has the following main properties:
1. The semantic field is intuitively understandable to a native speaker and has a psychological reality for him.
2. The semantic field is autonomous and can be singled out as an independent language subsystem.
3. The units of the semantic field are connected by certain systemic semantic relations.
4. Each semantic field is connected with other semantic fields of the language and together with them forms a language system.
The field stands out nucleus, which expresses the integral seme (archiseme) and organizes the rest around itself. For example, field - human body parts: head, hand, heart- the core, the rest are less important.
The theory of semantic fields is based on the idea of the existence of certain semantic groups in the language and the possibility of the occurrence of language units in one or more such groups. In particular, vocabulary language (lexicon) can be represented as a set of separate groups of words united by various relationships: synonymous (boast - brag), antonymous (speak - be silent), etc.
The elements of a separate semantic field are connected by regular and systemic relations, and, consequently, all the words of the field are mutually opposed to each other. Semantic fields may intersect or completely enter one into the other. The meaning of each word is most fully determined only if the meanings of other words from the same field are known.
A single linguistic unit can have several meanings and, therefore, can be assigned to different semantic fields. For example, the adjective red can be included in the semantic field of color designations and at the same time in the field, the units of which are united by the generalized meaning "revolutionary".
The simplest kind of semantic field is field of paradigmatic type, the units of which are lexemes belonging to the same part of speech and united by a common categorical seme in meaning, between units of such a field of connection of a paradigmatic type (synonymous, antonymic, genus-species, etc.). Such fields are often also called semantic classes or lexico-semantic groups. An example of a minimal semantic field of a paradigmatic type is a synonymous group, for example, the group verbs of speech. This field is formed by verbs talk, tell, talk, talk and others. The elements of the semantic field of verbs of speech are united by the integral semantic sign of "speaking", but their meaning not identical.
The lexical system is most fully and adequately reflected in the semantic field - the lexical category higher order. Semantic field - it is a hierarchical structure of a set of lexical units united by a common (invariant) meaning. Lexical units are included in a certain SP on the basis that they contain the archiseme that unites them. The field is characterized by a homogeneous conceptual content of its units; therefore, its elements are usually not words that correlate their meanings with different concepts, but lexico-semantic variants.
The entire vocabulary can be represented as a hierarchy of semantic fields of different ranks: large semantic spheres of vocabulary are divided into classes, classes into subclasses, etc., up to elementary semantic microfields. The elementary semantic microfield is lexico-semantic group(LSG) is a relatively closed series of lexical units of one part of speech, united by an archiseme of a more specific content and a hierarchically lower order than the archiseme of the field. The most important structuring relation of elements in the semantic field is hyponymy - its hierarchical system based on genus-species relations. Words corresponding to specific concepts act as hyponyms in relation to the word corresponding to the generic concept - their hypernym, and as cohyponyms in relation to each other.
The semantic field as such includes words different parts speech. Therefore, the units of the field are characterized not only by syntagmatic and paradigmatic, but also by associative-derivational relations. SP units can be included in all types of semantic categorical relations (hyponymy, synonymy, antonymy, conversion, derivational derivation, polysemy). Of course, not every word by its nature enters into any of these semantic relations. Despite the great diversity in the organization of semantic fields and the specifics of each of them, we can talk about a certain structure of the joint venture, which implies the presence of its core, center and periphery (“transfer” - the core, “donate, sell” - the center, “build, cleanse” - periphery).
The word appears in the SP in all its characteristic connections and various relationships that actually exist in the lexical system of the language.
Figure 2
Field types
Figure 1. Presentation of information in the database
Basic concepts
Database fields
The language of modern DBMS
The language of the modern DBMS includes subsets of commands that previously belonged to the following specialized languages:
Data description language - a high-level non-procedural language of a declarative type, designed to describe the logical structure of data.
Data Manipulation Language is a DBMS command language that provides basic operations for working with data - input, modification and selection of data by request.
Structured query language (Structured Query Language, SQL) - provides data manipulation and determination of the relational database schema, is a standard means of accessing the database server.
Ensuring the integrity of the database - necessary condition successful operation of the database. Database integrity is a property of a database, which means that the database contains complete and consistent information necessary and sufficient for the correct functioning of applications. Security is achieved in the DBMS by encryption of application programs, data, password protection, support for access levels to a separate table.
Field- the smallest named element of information stored in the database and considered as a whole.
The field can be represented by a number, letters, or a combination of them (text). For example, in a telephone directory, the fields are surname and initials, address, telephone number, i.e. three fields, all text fields (the phone number is also treated as some text).
Recording- a set of fields corresponding to one object. Thus, a subscriber of the telephone network corresponds to a record consisting of three fields.
File- a set of records related by some attribute (i.e. relation, table). Thus, in the simplest case, the database is a file.
All data in the database is divided by type. All field information belonging to the same column (domain) is of the same type. This approach allows the computer to organize the control of the input information.
Main types of database fields:
Symbolic (text). This field can store up to 256 characters by default.
Numerical. Contains numerical data in various formats used for calculations.
Date Time. Contains a date and time value.
Monetary. Includes monetary values and numeric data up to fifteen integer and four fractional digits.
Note field. It can contain up to 2^16 characters (2^16 = 65536).
Counter. A special numeric field in which the DBMS assigns a unique number to each record.
Logical. Can store one of two values: true or false.
OLE (Object Linking and Embedding) object field. This field can contain any spreadsheet object, microsoft word document, picture, sound recording, or other binary data embedded in or associated with the DBMS.
Substitution master. Creates a field that offers a choice of values from a list or containing a set of constant values.
Database fields do not just define the structure of the database - they also define the group properties of the data written to the cells belonging to each of the fields.
The main properties of database table fields are listed below using the Microsoft Access DBMS as an example:
Field name- determines how the data of this field should be accessed during automatic operations with the database (by default, field names are used as table column headings).
Field type- defines the type of data that can be contained in this field.
Field size- defines the maximum length (in characters) of data that can be placed in this field.
Field Format- determines how data is formatted in the cells belonging to the field.
input mask- defines the form in which data is entered in the field (data entry automation tool).
Signature- defines the table column heading for the given field (if the label is not specified, then the Field name property is used as the column heading).
Default value- the value that is entered into the field cells automatically (data entry automation tool).
Value condition- a constraint used to validate data entry (an entry automation tool that is typically used for data that has a numeric, currency, or date type).
Error message - text message, which is issued automatically when you try to enter erroneous data in the field (error checking is performed automatically if the Condition on value property is set).
Required field- a property that determines the mandatory filling of this field when filling the database.
Blank lines- a property that allows the input of empty string data (it differs from the Required field property in that it does not apply to all data types, but only to some, for example, text).
Indexed field- if the field has this property, all operations related to searching or sorting records by the value stored in this field are significantly accelerated. In addition, for indexed fields, you can make it so that the values in the records will be checked against this field for duplicates, which automatically eliminates data duplication.
Since different fields may contain data of different types, the properties of the fields may differ depending on the type of data. So, for example, the list of field properties above applies primarily to fields of the text type. Fields of other types may or may not have these properties, but may add their own to them. For example, for data representing real numbers, important property is the number of digits after the decimal point. On the other hand, for fields used to store pictures, sound recordings, video clips, and other OLE objects, most of the above properties are meaningless.
semantic field
semantic field
The semantic field is a set of words united by semantic connections according to similar features of their lexical meanings.
In English: semantic field
See also: Languages
Finam Financial Dictionary.
See what "Semantic field" is in other dictionaries:
SEMANTIC FIELD- SEMANTIC FIELD. A set of words and expressions that make up a thematic series that is stored in a person's long-term memory and occurs whenever communication is necessary in a certain area. Creation of S. p. in human memory - ... ... New dictionary methodological terms and concepts (theory and practice of teaching languages)
SEMANTIC FIELD- see Semantics. Big psychological dictionary. Moscow: Prime EUROZNAK. Ed. B.G. Meshcheryakova, acad. V.P. Zinchenko. 2003 ... Great Psychological Encyclopedia
1) A set of phenomena or an area of reality that has a correspondence in the language in the form of a thematically united set of lexical units. The semantic field of time, the semantic field of space, the semantic field of souls ... ... Dictionary linguistic terms
The same as the lexico-semantic field...
The largest semantic paradigm that unites words of different parts of speech, the meanings of which have one common semantic feature. For example: SP light light, flash, lightning, shine, sparkle, light, bright, etc. Contents 1 Dominant 1.1 Fields ... Wikipedia
semantic field- an extensive association of words related in meaning, determining and predetermining the meanings of each other. S. P. reflects the connections and dependencies between the elements of reality, objects, processes, properties, therefore it naturally includes ... ... Russian humanitarian encyclopedic dictionary
semantic field- 1. A set of words and expressions that make up a thematic series; words and expressions of the language, in their totality covering a certain area of knowledge. 2. A group of words whose meanings have a common semantic component. 3. The totality of phenomena ... ... Explanatory Translation Dictionary
semantic field- The largest lexico-semantic paradigm that combines words of different parts of speech, correlated with one fragment of reality and having a common feature (common seme) in lexical meaning … Dictionary of linguistic terms T.V. Foal
A set of lexemes denoting a certain concept in the broad sense of the word: modern ideas, the field includes in its composition the words of various parts of speech, with the assumption of the inclusion of phraseological units and lexical materials of various ... ... Handbook of etymology and historical lexicology
Term functional grammar; based on a certain semantic category, a grouping of means of different levels of the language, as well as combined linguistic means interacting on the basis of their commonality semantic functions. This ... ... Wikipedia
Books
- Semantics of information aspects, L. A. Kochubeeva, V. V. Mironov, M. L. Stoyalova. The book presents the results of a three-year study of the St. Petersburg working group. Experimentally verified and systematized data that representatives of different socionic ...
- From A to Z. The most complete encyclopedia of aphorisms, thoughts and quotes, Polyakov Yuri Mikhailovich. The book is the most complete collection of aphorisms, thoughts and quotes to date, extracted from prose, poetry, plays, journalism, interviews and notebooks of the famous ...
Random fields are called random features many variables. In the future, four variables will be considered: coordinates, which determine the position of a point in space, and time. The random field will be denoted as . Random fields can be scalar (one-dimensional) and vector (-dimensional).
In the general case, a scalar field is given by the set of its -dimensional distributions
and the vector field - a set of its own - dimensional distributions
If a statistical characteristics fields do not change when the time reference changes, i.e. they depend only on the difference, then such a field is called stationary. If the transfer of the origin does not affect the statistical characteristics of the field, i.e., they depend only on the difference, then such a field is called spatially homogeneous. A homogeneous field is isotropic if its statistical characteristics do not change when the direction of the vector changes, i.e., they depend only on the length of this vector.
Examples of random fields are the electromagnetic field during the propagation of an electromagnetic wave in a statistically inhomogeneous medium, in particular, the electromagnetic field of a signal reflected from a fluctuating target (generally speaking, this is a vector random field); volumetric radiation patterns of antennas and patterns of secondary radiation of targets, the formation of which is influenced by random parameters; statistically uneven surfaces, in particular earth's surface and the surface of the sea during waves, and a number of other examples.
In this section, some issues of modeling random fields on a computer are considered. As before, the modeling task is understood as the development of algorithms for the formation of discrete field realizations on a digital computer, i.e., sets of sample values of the field
,
where - discrete spatial coordinate; - discrete time.
In this case, it is assumed that independent random numbers are the initial ones when modeling a random field. The set of such numbers will be considered as a random -correlated field, hereinafter called -field. A random -field is an elementary generalization of discrete, white noise to the case of several variables. Modeling of the -field on a digital computer is carried out very simply: the space-time coordinate is assigned a sample value of a number from a generator of normal random numbers with parameters (0, 1).
The problem of digital simulation of random fields is new in the general problem of developing a system of efficient algorithms for simulating various kinds of random functions, focused on solving statistical problems of radio engineering, radiophysics, acoustics, etc. by computer simulation.
In the very general view, if the or -dimensional distribution law is known, a random field can be modeled on a computer as a random or -dimensional vector using the algorithms given in the first chapter. However, it is clear that this path, even with a relatively small number of discrete points along each coordinate, is very complicated. For example, the simulation of a flat (independent of ) scalar random field at 10 discrete points along coordinates and and for 10 time moments is reduced to the formation on a computer of realizations of a -dimensional random vector.
Simplification of the algorithm and reduction in the volume of calculations can be achieved if, similarly to what was done with respect to random processes, algorithms are developed for modeling special classes of random fields.
Consider possible algorithms for modeling stationary homogeneous scalar normal random fields. Random fields of this class, like stationary normal random processes, play a very important role in applications. Such fields are completely specified by their spatiotemporal correlation functions
(Here and in what follows, it is assumed that the mean value of the field is zero.)
An equally complete characteristic of the considered class of random fields is the field spectral density function , which is a four-dimensional Fourier transform of the correlation function (a generalization of the Wiener-Khinchin theorem):
,
where is the scalar product of the vectors and . Wherein
.
The spectral density function of a random field and the energy spectrum of a stationary random process have a similar meaning, namely: if a random field is represented as a superposition of space-time harmonics with a continuous frequency spectrum, then their intensity (total amplitude dispersion) in the frequency band and spatial frequency band is equal to .
A random field with intensity can be obtained from a random field with spectral density , if the field is passed through a space-time filter with a transfer coefficient equal to unity in the band , and equal to zero outside this band.
Spatio-temporal filters (SPFs) are a generalization of conventional (temporal) filters. Linear PVFs, like ordinary filters, are described using the impulse response
and transfer function
.
The process of linear space-time field filtering can be written as a four-dimensional convolution:
(2.140)
where is the field at the output of the PVF with an impulse transient response. Wherein
where are the spectral density functions and the correlation functions of the fields at the input and output of the PVF, respectively.
The proof of relations (2.141), (2.142) completely coincides with the proofs of similar relations for stationary random processes.
The analogy of harmonic expansion and filtering of random fields with harmonic expansion and filtering of random processes allows us to propose similar algorithms for their modeling.
Let it be required to construct algorithms for computer simulation of a stationary, space-homogeneous scalar normal field with a given correlation function or spectral density function .
If the field is given in a finite space, bounded by the limits , and is considered on a finite time interval , then to form discrete realizations of this field on a computer, one can use an algorithm based on the canonical expansion of the field in the space-time Fourier series and which is a generalization of algorithm (1.31):
Here, and are random mutually independent normally distributed numbers with parameters each, and the variances are determined from the relations:
where is a vector representing the limit of integration over space; - discrete frequencies of harmonics, according to which the canonical expansion of the correlation function is performed in the space-time Fourier series.
If the field expansion area is many times larger than its spatiotemporal correlation interval, then the dispersions are easily expressed in terms of the field spectral function (see § 1.6, item 3)
The formation of discrete realizations when modeling random fields using this method is carried out by directly calculating their values according to (formula (2.143), in which sample values of normal random numbers with parameters are taken as and , while the infinite series (2.143) is approximately replaced by a truncated series. Variances are calculated previously by formulas (2.144) or (2.146).
Although the algorithm considered does not allow one to form realizations of a random field that are unlimited in space and time, the preparatory work for obtaining it is quite simple, especially when using formulas (2.145), and this algorithm allows one to form discrete field values at arbitrary points in space and time selected area. When forming discrete realizations of a field with a constant step along one or several coordinates for an abbreviated calculation trigonometric functions it is expedient to use a recurrent algorithm of the form (1.3).
Unlimited discrete implementations of a homogeneous stationary random field can be formed using space-time sliding summation algorithms -fields, similar to sliding summation algorithms for modeling random processes. If is the impulse transient response of the PVF, which forms a field with a given spectral density function from the -field (the function can be obtained by four-dimensional Fourier transform of the function , see § 2.2, item 2), then, subjecting the process of spatiotemporal filtering of the -field to discretization, we get
where - a constant determined by the choice of the sampling step over all variables - discrete -field.
The summation in formula (2.146) is carried out over all values for which the terms are not negligible or equal to zero.
The preparatory work for this modeling method is to find the appropriate weight function of the space-time shaping filter.
The preparatory work and the summation process in the algorithm (2.146) are simplified if the function can be represented as a product
In this case, as follows from (2.144), the correlation function of the field is a product of the form
If the decomposition of the correlation function into factors of the form (2.148) is not feasible in the strict sense, it can be done with a certain degree of approximation, in particular, by setting
When decomposing into a product (2.149) of spatial, correlation functions of isotropic random fields, for which , partial correlation functions and will obviously be the same. In this case, in view of the approximation of formula (2.149), the spatial correlation function will correspond, generally speaking, to some non-isotropic random field. So, for example, if is an exponential function of the form
then according to (2.149) . In this case, the given correlation function is approximated by the correlation function
. (2.151)
The random field with the correlation function (2.151) is not isotropic. Indeed, if a field with correlation function (2.150) has a constant correlation surface (the locus of space points where field values have the same correlation with the field value at some arbitrary fixed point in space) is a sphere, then in case (2.151) the constant correlation surface is the surface of a cube inscribed in a given sphere. (The maximum distance between these surfaces can serve as a measure of the approximation error).
An example in which expansion (2.149) is exact is a correlation function of the form
Decomposition (2.149) allows us to reduce the rather complicated process of quadruple summation in algorithm (2.146) to the repeated application of a single sliding summation.
These are the basic principles of modeling normal homogeneous stationary random fields. Modeling of non-normal homogeneous stationary fields with a given one-dimensional distribution law can be done by an appropriate non-linear transformation of normal homogeneous stationary fields using the methods discussed in § 2.7.
Example 1 Let the impulse response of the spatial filter for the formation of a flat scalar time-constant field have the form
where and are discretization steps in variables and with a weight function form discrete realizations of the field. The process of such double smoothing - the field is illustrated in Fig. 2.11.
In the example under consideration, the process of moving summation can easily be reduced to a calculation in accordance with the recursive formulas (§ 2.3)
This example allows for generalizations. First, in a similar way, it is obviously possible to form realizations of more complex fields than a flat, time-constant field. Secondly, the example suggests the possibility of using recurrent algorithms for modeling random fields. Indeed, if the impulse transient response of the PVF, which forms a field with a given correlation function from the -field, is represented as a product of the form (2.151), then, as was shown, the formation of field realizations is reduced to the repeated application of algorithms for modeling stationary random processes with correlation functions . These algorithms can be made recurrent if the correlation functions , have the form (2.50) (stochastic processes with rational spectrum).
In conclusion, it should be noted that in this section only the basic principles of digital modeling of random fields have been considered and some possible modeling algorithms have been given. A number of issues remained untouched, for example: modeling of vector (in particular, complex), non-stationary, non-homogeneous, non-normal random fields; questions of finding the weight function of the space-time shaping filter according to the given correlation-spectral characteristics of the field (in particular, the possibility of using the factorization method for multidimensional spectral functions); examples of the use of digital models of random fields in solving specific problems, etc.
The presentation of these questions is beyond the scope of this book. Many of them are the subject of future research.
temperature field- a set of temperature values at all points of the body at a given time. Mathematically, it is described as
where x, y, z- spatial coordinates;
t- the time of the thermal process.
There are two characteristic cases of the temperature state of the body:
1. At each point of the body, the temperature remains unchanged in time, i.e.
In this case, the temperature at different points of the body can be the same or different. The temperature state of the body, unchanged in time, is called stationary (steady). In this state of the body, the heat input is equal to its consumption.
In a stationary thermal regime, masonry works blast furnace, continuous thermal and heating furnaces, recuperators. The heating time of the furnace to the operating temperature in these devices is negligible compared to the operating time of the furnace at a given temperature.
2. When a body is heated or cooled, the temperature at each point of it continuously changes in time. Such a temperature state of the body, in which the temperature is a function of both coordinates and time, is called non-stationary (unsteady). In this mode, the laying of batch furnaces (bogie hearth furnaces, heating wells, open-hearth furnaces), as well as the packing of regenerators, works.
If the body temperature changes only along one spatial coordinate, the temperature field is called one-dimensional.
temperature gradient- the limit of the ratio of the temperature increment between two isotherms to the distance between them, measured along the normal.
(37)
heat flow- the amount of heat transferred per unit of time ( Q, W) across the entire surface.
Vector grad t is considered positive if it is directed in the direction of increasing temperature, and the heat flux vector Q is positive if it is in the direction of decreasing temperature.
If the heat flux is attributed to the surface unit, then we obtain the heat flux density, W/m 2 .