CC BY
22
COMPUTER SCIENCE
MODELING OF THE MACHINE TRANSLATION SYSTEM FOR TECHNICAL TEXTS OF THE TELECOMMUNICATION FIELD USING THE OBJECT-ORIENTED PROGRAMMING LANGUAGE C #
Andrey V. Alyoshintsev,
Moscow Technical University of communication and informatics, Moscow, Russia, [email protected]
Elena V. Bessonova,
Moscow State University of Civil Engineering, Moscow, Russia, [email protected]
A|exander N. Sa^ Keywords: machine translation, object-oriented
Moscow State University of Civil Engineering, programming, lexeme, grammar, arrays,
Moscow, Russia, [email protected] collections, online dictionaries.
The need for constant access to scientific articles or instructions in English on the use of devices like high frequency modems for instance, as well as the emergence of new technologies, requires the development of a machine translation system for the telecommunication industry. The development of object-oriented programming languages enables to represent objects containing a large amount of information in the most convenient way from the point of view of the tasks faced by a developer. The article deals with the presentation of lexical units of the sentence as objects of the program in C # language, as well as the peculiarities of translating the corresponding lexemes, taking into account the grammar of English and Russian in the field of telecommunications. The authors try to avoid the presentation of the lexical-semantic structure of the sentence by means of a metalanguage as a mediator, which significantly complicates the translation system, by creating a lighter and more understandable structure. For this, certain rules are established for the lexical and grammatical relationship of the narrative sentence in English and Russian. The peculiarities of processing html-pages of online dictionaries used as a thesaurus of lexical units in translation are taken into account.
Information about authors:
Andrey V. Alyoshintsev, Moscow Technical University of communication and informatics, Moscow, Russia Elena V. Bessonova, Moscow State University of Civil Engineering, Moscow, Russia Alexander N. Sak, Moscow State University of Civil Engineering, Moscow, Russia
Для цитирования:
Алешинцев А.В., Бессонова Е.В., Сак А.Н. Моделирование системы машинного перевода технических текстов в телекоммуникационной сфере с использованием объектно-ориентированного языка программирования С# // T-Comm: Телекоммуникации и транспорт. 2017. Том 11. №10. С. 66-73.
For citation:
Alyoshintsev A.V., Bessonova E.V., Sak A.N. (2017). Modeling of the machine translation system for technical texts of the telecommunication field using the object-oriented programming language C #. T-Comm, vol. 11, no.10, pр. 66-73.
7ТЛ
The machine translation system for the telecommunication area can be regarded as an information system improving the organization of telecommunication logistics.
The problem of machine translation is located at the junction of linguistics and cybernetics and is one of the tasks to address when studying artificial intelligence problems.
At present, when developing systems in the area of telecommunications, the complex systems of data processing and analysis increasingly comprehend subsystems for processing textual information. If such subsystems are designed to work with data in several languages, they may face the task of automatically translating from one language to another. Existing machine translation systems make mistakes when translating texts from the telecommunication field. For example, the Google machine translation system translates the phrase "Any action that must be taken whether or not an exception is raised should be enclosed in a finally block" [l| as «Любое действие, которое должно быть предпринято независимо от того, возбуждено ли исключение, должно быть заключено в блок finally». As we can see, instead of Russian verb "запускать - to launch", the verb "возбуждать-to excite" is used. The development of automation tools for translation in the communication area is useful even with small volumes of translated text, since in this case the human factor is excluded: the need to contact a professional translator every time an engineer needs to translate a text saturated with unfamiliar lexical and grammatical units is eliminated. The use of online dictionaries of the relevant technical field will allow you to get the most adequate translation. Machine translation systems, based only on statistical algorithms, experience great difficulties in translating phraseological expressions and idioms as weil as the slightest deviation from simple grammatical constructions. Systems based on linguistic rules have difficulty translating technical terms. Our system, using online dictionaries, both bilingual and
monolingual, either general or specialized, is aimed at solving these contradictions.
Object-oriented programming languages, such as C #, expect the creators of machine translation systems to find new approaches to the implementation of these systems, based on the capabilities of these languages and the search for those linguistic methods that are compatible with the capabilities of modern programming.
Our system uses the object-oriented programming language C # for these purposes. Due to the fact that C # is a hybrid made out from several languages, it is syntactically pure, like Java, simple as VB, powerful and flexible, like C ++. The choice of this language for work is determined by the simplicity of creating classes and the ease of operations with their objects. C # has several advantages such as: it's easy to create collections of both string and those containing class objects, to convert quickly collections to arrays, to control the reliability of array elements, as well as to support optional parameters and named arguments in methods.
At the first stage of the analysis, a query is made about each lexical unit of the sentence to the online monolingual explanatory dictionary. It is important to note that the use of online dictionaries of narrow specialization in the field of telecommunications enables to achieve the most adequate translation in each of the corresponding directions. Possibilities of using words as different parts of speech are reflected in the corresponding collections included in a class object that characterizes given lexical unit. Having collections of uses and properties of each lexeme of the sentence, we can identity each word from the point of view of its belonging to one or another part of speech, which is very important because to solve the problem of polysemy or homonymy it is necessary to determine which part of speech the lexeme belongs to in this context [2].
гй r^U./ij ', viwel iiudiu
i<-" Ernprt guild firtm? Тещ loot*
^ A J t. -J Л 1 - • »!• *
^ k A* Tt i* -j ■ ■ . *! 4
P.DC*^ (HTjJ nlmll ПС
в
IE
XJ diplcmaji.Foiml
»I ./»button]" ,C!«l((objMt sender. EventArjt el
//д**е p случм нахождение подобие re м>р4*еи*в-пр*дгтск сдемтъ мг н«ид н проверить «j string teepö ■ itrlnj tenpl ■ "*t v ■ int cti ■ в; int kSS • 1; int itop - I; for (int i • i < suped, length; 1+4)
< P Прсдло»«**«
Hf (иртгЩТ !- nullH { * supel (iüplom«33/ofml,WorcWeMurts(lO)l
H (üni / ¡0! lfcptoira35.Forml.WoftWe«itf«| { V Dl (Äptoma33-Fcrml.Wordfeatww I
1 . adjectivel |stnng[2000]l
• ■ у adverb! jitring[2000]] f
< ♦ V article! {«гмдДОЦ]
* j fl|fl*wi) A » 4ff4irtri!«n*
A aw »urti lHfing(2000]J
,ij IrttefciTrace debugging о available only for «S6 applications- То enabt* InteftTrace debugging change the perform 1c
Open hrtefliTrace !irt!mg: friable Just My Cod«
* P3 A - "rtöurt"
4 [11 A * ЧЬе «t of toncenOafiny the rtate d b«ng concentrated"
- [21 4 * "CKluuve attention to one object' do» mental application"
4 [31 Л • "something concenlnled: a tone в*лоо of Sin1
* Ю Л - "Military'
W [5] 4 • 'Ibe assembling of military or nival force; In в particular area in preparation for further operations'
* [6] 4 - "a specified intensity and duration of a*tt*ery fire placed on a small area"
;*m 4 • "the focusing of a student's aeade htii< program on ach-anCed slutfy in a spei ific subject or add"
. 4 [3] я * " Chenuitry. (in a solution) > meas vre of the »nmuD of dissolved substance ccnü«ed per of volume'
U [9] a - "Abo tailed memwy, Cardi a gai ne in which all 52 card! не spread out face down on the table and each pityin turn expos« two cwds al a time wd repliiei them face doMti rf they do not сonartute a pa*f,...
4 [10} A * 'OriginEipamT
fill] * * "concert^«} * -rtion"
* tui Л • " Related fo<ms£icpand~
* [131 ч - 'hyp«fcon<entrjtioa noun"
V [Щ а • "тклсопселиаПюп. noun'
El Autoi Ц Lcxal Ready % Сав Stack Ü Вгеа*ро«пЬ Ш C< »mnund Window * [mmedute Window Bil'TiTI РгД'.ШИ ^ SoL ^ Tea,. Ln 3S2 Cot 16 Ch 16 IMS
© © о« « k^: (V Q Л ее w iL
Fig. 1. Assignment of used program variables
f I Л
Using classes and their objects when working with lexemes allows each lexeme to be represented in the most compact form, which facilitates its subsequent processing. This approach allows referring effectively to lexical units* features that are needed at this stage of the translation.
In fact, an automaton which is used recognizes different elements of the set A, where A is the collection of words
contained in the language. The automaton reads and recognizes the a0ala2a ... .an line if it stops in the final state after reading, that is, it determines the belonging to the part of speech of each word of the sentence, and at this final stage of analysis all parts of speech by their elements end up by composing a logical structure.
Z Parsing sentence / into words /
| Identifying other members as parts of speech from the i-.th element from the right to Ulu4«JEt-^-
If i-th element is a noun
Mo
yes
Searc fnr VP liing rbi
yes
Searchingfor su bject
yes
.The eventual subject and predicate have beer found
yes
Id entifyi ng oth e r s e n t e n ce members as pa rts of speech from the right to the left ¡beginning with an i-th ef. ,v'_rb;
Searching for idioms and sayings
No
Searching for phrasal verbs
If the word is the fast one in the array
identifying other Yes If it's a noun If it's a
sentence members preposition
Bs parts of speech
from the rieht to left Nc
Identifying other sentence members as parts of speech
Fig. 2. Block diagram of the algorithm for identifying parts of speech in a sentence
When the words of a sentence are distributed by parts of speech, i.e. we can say that the process of the sentence analysis in our study is completed, it is necessary to proceed to the adaptation of the received lexemes to Russian grammar. Comparing the narrative (as the commonest in the translation of scientific and in particular texts on communication) sentences of Russian and English, one can say that in principle their grammatical structure is similar: the subject-predicate-the direct compIement-the indirect complement-the circumstance of the place-the circumstance of time. Just like in English, the order of
sentence members in the Russian narrative sentence can vary within the framework of certain rules, which gives sufficient room for maneuver [3]. The main difference is the fact that nouns, adjectives and participles are changed according to cases in Russian, in English the relations between members of the sentence are carried out mainly by means of prepositions [4]. Focusing on the importance of verbs and prepositional constructions in a phrase, a certain set of rules is developed, allowing to build the grammatical structure of a Russian sentence by analyzing the vertex of the English sentence-the verb and its
linguistic environment. The refusal to use the intermediate-language (a met a-language) can be considered a method that facilitates the translation of this type of technical text.
When implementing a system of translation from English into Russian, it is necessary to take into account the prepositional system of English verbs arrangement.
The idea of the synthesis process is to decide on the linguistic environment in which form the appropriate word or construction in the target language will be used, and the use of matrices greatly facilitates this task.
The authors adapt the grammatical pointers used in the program to the pointers adopted in the dictionary, taking into account the peculiarities of the online English-Russian dictionary.
prédicat«
leave
subject
tomorrow
cireonstante
Fig. 3. An example of distribution of actants in a sentence
core.rling.First;
texteoxi.Text += core.transl+'\r\n"+
//textBojrt.Text core.transz; Li nk ed ListNode(interpreter» nodeiee k = e; string ssss = "; string raxeee -for (irvt xyt m 0; xyz t S; xy2++)
// texteoxl.Text sentencemenbersll, xyz] texteoxl.Text "\r\n" + "this is a pattern" * int max _ -1000; string ssmax string sssm
int nttnaxl = -1800; string ssmaxl = ""; string ss int mrnin - ie«0; string ssntin « string sssmi int mmini m 1000; string ssmini = ""; string sss
while (nodeiee != null) {
string[] words = nodeiee.value.Getengiish(). for (int sad=e;sad<wrds.Length;sad++)
if (words[sad] — sentencememberste, poi
i
string[] wordsl new string [words. L
//words.CopyTo(wordsl, sad);
for (int pop = e; pop < words.Length
{
wordsl[pop] « words[sad + pop];
>
syntax current » new syntax(wordsl, sentencemem
wordfeature:- templll;
LinkedList<wordfeatures> nodel23 ■ new Lir
//nodeizJ. AddLsst(current. &e tSuper:(e));
-\r\n-
a / core.rlinq Counts 102
✓ [0] {neiArtranslatjon.interpreter!
English -1 -г "leave"
ж
Ф [0] * * "покрываться листвой, "
у [11 » "покрыться листвой "
у [2] ▼ "покидать, "
v [3] А * "покинуть "
Ф [4) null
/ [5] null
[6] null
V [7] null
ф [8] null
ф [9] null
Ф [Ю] null
У [11] null
/ [12] null
• [13] null
у [14] null
*
Fig. 4, Part of the program showing an online bilingual dictionary at work regarding the English verb "to leave"
and some options of its translation into Russian
core,transiatcr(ref sentencememtersie, i], ref sentencemembers[i, i]j supers, prepiist);
..............................................-v\r\n";
texteoxl.Text +- core.transl+"\r\n"+".......
//textBoxl.Text core,trans2; LinkedListtoJe*interpreter> nodeiee = core.rlino. к ж в; string ssss « ""; string иахаее = si a for (int xyz - 9; хуг < 5; xyz++)
// texteoxl.Text serrteneemeaberstl, xy2] + textBoxl.Text +- "\r\n" + "this is a pattern" + ' int max * -ieee; string ssmax » "j string sssma int ranaxl = -leeej string ssmaxl = ""; string sss int imin = ieee; string ssmin я ""; string sssmir int mini • lⅇ string ssmini = **; string sssif
while (nodeiee !<■ null)
{
stringt] words = nodeiee.value.GetEngiish().s for (int sad=e;sad<words.Length;sad++)
if (words[sad] « seritencememoerste, poir
{
string[] wordsl = new string[words.Li
//words.соруто(wordsl, sad);
for (int pop = e; pop < words.Length
first;
/ core.rlinq Count = 102
Qfl Ф [0] a * Ml
{newtranslation.interpret er)
a / [2] {newtranslation.interpreter)
El engiish i - "my sister has left for moscow"
m a j/ russianl {string! 15001)
a V Ж
a Ф * Ю1 '-I » "сестра уехала в москву "
a Ф ♦ Ml » "оставлять в том же состоянии, "
a ф « [21 4. - "оставить е том же состоянии, *
a Ф ♦ [31 ^ - "не касаться "
a ф ♦ [41 null
a ф ♦ [51 null
a ф ♦ [61 null
m y ♦ 171 null
Fig. 5. Other variants of translation of English verb "to leave" in the online bilingual dictionary
7ТЛ
if we used a monolingual online dictionary when identifying parts of speech, then at the second stage we use the English-Russian online dictionaiy "English-Russian translation of words and expressions, definition, synonyms // Re verso-Dictionary". This dictionary takes into account a lot of verb usages and their prepositional control, mentioning each case of such use. For example, the sentence "1 leave Moscow for Paris tomorrow" can be represented by the following graph, in which additions to the predicate occupying a grammatically prescribed position in accordance with the semantic load are called "actants" - the obligatory elements of the actual division of the sentence in connection with its Taxonomic role:
Obviously, the dictionary proposes to translate Actantl as a direct complement using the verb "покидать" or "leave."
The Actant2 "for Paris" implies the use of the verb "io leave for...". Thus, in our case it is necessary to resolve this contradiction by means of the verb "to leave ... for ...", which combines the use of both actants or at least, translates as "leaving Moscow and leaving for Paris" if the use of "leaving something for something" was not found.
It is also important to note that special processing of the phrasal verb or phraseological expression is provided, for which another cycle is started, which collects all the elements of this stable expression into a single variable, which is then transferred for translation. From the point of view of the "Meaning О Text" model, a transition from the surface-syntactic structure to the deep-syntactic structure is observed here.
English adjectives stand before nouns as in Russian. Nouns are often used just as adjectives in English: data hiding -"сокрытие данных". The word "data" is a noun acquiring properties of an adjcctive here, and the full nominalization "hiding" is a noun that comes from the verb "to hide-прятать, срывать". So that is translated into Russian by means of two nouns, where the determiner is supposed to be in the possessive case and to follow the main word.
It is also necessary to talk about the technology of generation and transmission of characteristics for processing words already received as a result of translation. The matrix "sentenc em embers" contains detailed information about each word of the sentence. The Oth line contains the English word itself, the 1st line - contains data about what part of speech it corresponds to, if this is a verb, then in the second line it is indicated depending on the detected predicate the singular or the plural form of the verb and to which person its shape corresponds to. If it is a noun, then in the second line it will be indicated whether singular or plural it is used in. The meaning of the synthesis process is to decide, according to the linguistic environment, in which form the corresponding word or construction in the target language will be used.
[f, for example, there are two nouns together, between which there is a preposition "of', then, most likely, there is a possessive relation between words and in Russian the second name should be put in the possessive case and it is natural to forget about the preposition. It is stipulated not to translate prepositions by means of a dictionary, but to provide the program with a way of translation. In this case, it is necessary to remove the preposition and put the noun following the preposition in the possessive case. The analysis is sequential, i.e. done word by word, and the matrix " sentenceniembersi" is formed, where the Russian words will be placed after the translation, and their characteristics will
be determined in advance by the linguistic environment. Therefore, the characteristic of the word that should be in the possessive case will be transferred to the next element of the array
if (chl - 1 >= 0 && cht + 1 <= super2.Count())
{
if (sentencemembers[l,ehl-l]="noun" &&
sentencemembers[ i ,ch 1 +1 ]="noun")
sentenceniembersi [3,chl + l]="genetive";
}
And the translation of the next element will happen already in the next step of the cycle.
If there is a combination "to" + verb, then the preposition "to" must be replaced with "чтобы", directly, without referring to a dictionary. If the preposition "to" is followed by a noun, an adjective, or an article, there are several options for translation. This can be translated with prepositions like "в", "на" or "к", depending on the context, and if the preposition is followed by a noun, it must have the dative case in Russian. If we are talking about the sequence "the preposition "in" + noun", we translate the English preposition into Russian by the preposition "в", and the noun is put into the prepositional case, and if after the preposition there is an adjective, once the translation is done, the adjective should also have the form of a prepositional case. The same principles are extended to prepositions as "for" then the noun following it will be in the objective case, the preposition "from" corresponds to the preposition "из" + the noun in the possessive case, the preposition "about" corresponds to the preposition "насчет" and a noun in the possessive case, the preposition "without" with a name corresponds to the preposition "без" + noun the possessive case, and if the preposition is used with a verb in the form of a gerund, it must be translated using the gerundive turnover, for example, "without doing anything." If it is a preposition "with", then it must be translated using the preposition "c" + noun in the instrumental case.
Taking into account the peculiarities of the online English-Russian dictionary, one should adapt the pointers of the parts of speech to the corresponding pointers of the dictionary: "noun" —* "n", "adjective" —+ "adj", "pronoun" —» "pron", in case of an adverb, then according to our observations, the "adj" marker should be indicated. In this dictionary there are two markers of the part of speech for the verbs "vt" and "vi", i.e. a transitive and non-transitive verb. For example, the verb "to run" in a nontransitive sense means "to run," and in a transitive one, for example, "to run an office" means "to manage an office". In order to find out which verb we are dealing with, we need to check whether there is a direct complement to this verb.
if (sentencemembers[l, chl] = "verb") {
str33 = "vi";
for(intx) = chl;xl <super2.Count();xl++) {
if (sentencemembers[l, xl] — "noun" ||
sentencemembers[l, xl] == "pronoun") {
str33 = "vt"; break; }
if (sentencemembersjl, xl] = "preposition")
break; }
If a phrasal verb or a phraseological expression occurs at this stage of the cycle, it's necessary to run another cycle that collects all the elements of this persistent expression into a single variable, which is then passed to the translation. From the point of view of the "Meaning »Text" model, a transition from a surface-syntactic structure to a deep-syntactic structure is observed here [3].
if (sentencemembers[l, chl ] = "phrasal verb") {
str25 +="";
for (int ch2 = chl + 1; ch2 < super2.Count(); ch2++) {
if (sentencemembers[l, ch2] — "phrasal verb") { str25 += sentencemembers[0, ch2] + " "; sentencememberslQ, ch21 = ""; 1
Since it is impossible to assign a corresponding token to a phrasal verb or phraseological expression, which would be perceived by the dictionary, an "empty" value is assigned to the variable str33 = "";
Some English expressions and constructions that are perfectly perceived by an explanatory dictionary are not available for this English-Russian dictionary and one has to produce a direct lexical setting:
if (str25 = "have to " || str25 = "has to ") str25 = "have got to";
And, finally, when the information on the translated word is prepared, the word is transmitted for translation
if (sentenccmembersfO, ch 1 ] != "the" && sentencemembers[0, chl] != "a" && sentencemembers[0, chl| != "" && sentencemembersj l,cM3l—"preposition11 &&
sentencemembers[0,chl]l="be") {
inst. trans lator(ref str2 5, ref str3 3); rlinq. AddLast(inst.node);
After processing the word by means of the class that handles the translation, the result is passed back to the Form using a reference to the object of the class, in whose role the collection of Russian values of the given word appears [5],
sentencem embers 1[0, chl] = inst. node. GetRussian(O); }
This block is focused on the way such complex prepositions as "in spite of or "instead of are handled. Bach element of
complex prepositions was introduced into an array of prepositions and when identifying parts of speech was determined as preposition. Then, in the above block, a cycle is started that analyzes whether this is an easy preposition or a complex one, and if the next word also has a preposition marker, then it is a complex preposition, the parts of which arc assembled and translated. There is a special class "translation", directly working with the English-Russian dictionary.
The method "translator" deals with the arguments of the word in English, pre-processed in the two previous blocks, as well as the part of speech to which it belongs.
Returning an already processed string, we continue processing it. Then, if a part of the speech is determined, the search goes on until the first letter in Russian is found. If such a letter is found, the sentence is broken up into words and words arc entered into the array. Looking through its elements, if there is an element identical to the definition of an already determined part of speech, there is a transition to the next element. If this element is in Russian, it is written down into the "tempi" variable and an "tempO" element is added to the English collection, and if the element is in English and provided that tempi! = null, we'll put the value of the tempi variable into the Russian collection and add the element to the "tempO" variable. In case of phraseology or a phrase verb, the first thing to do is to find a match of the line with the expression in English, which was obtained in the previous grammar block, and only then we represent the sentence words as array elements, which are then written down as a tempo and tempi variables are sent to collections named "English" and "Russian" correspondingly. After that, a given word, expressed by these collections, is represented as an instance of the class "interpreter". In this class, the values of the object "node" in English in different expressions and the values of the same object in Russian are represented by the arrays" englishl" and Lrussianl". This allows to coordinate English and Russian values at the expense of the index.
In this block, the stage of the pre-translational preparation of the sentence before making a request for a bilingual English-Russian dictionary was considered.
On the next stage of machine translation it will be necessary to process the received Russian lexemes from the point of view of Russian grammar.
In conclusion, we can say that a system that uses online dictionaries, rather than body of text as a thesaurus, should combine both statistical methods in the first stage of the proposal processing and rules at the level of deeper semantic analysis of the sentence. For more detailed semantic analysis, it is necessary to use, in addition to the taxonomic role and category of participants in the proposal, also their subject area and referential status.
It is necessary to give an account of the ideology of such a programming language as C # in order to create an approach to the machine translation system that would use its capabilities as efficiently as possible.
str33 =""; }
}
Fig. 6. Interface Form I with the implementation of an example of translation
References
1. Sannikov V,G,, Alyoshintsev A.V. (2015), Multi frequancy modem as one of the basic elements of the "smart building" system with a remote control of the objects [Mnogochastotnyi modem kak odin iz osnovnih elementOV si stem i "intellectualnoye zdaniye" pri udalennom upravlenii objectami]. T-Comm. No. 6. Pp 2!-27(/h Russian)
2. Sak A.N. (2015). Identification of the sentence members as the main task of machine translation [Identificacia chlcnov prcdlojenia kak osnovnaia zadacha mashinnogo perevoda]. Scientific Review, Ed.14. (in Russian).
3. Mel'Suk l.A. (2012). From the sense to the text [Ot smysla k tekstu] Moscow: Languages of Slavic culture, (in Russian).
4. Mel'Cuk I.A. (1995). Russian language in the model, "Meaning w1 Text" [Russkii yazyk v inodeli "Smysl-Tekst". Moscow - Vienna: School "Russian Culture Languages" Viennese Slavonic Almanac. XXV111, 682 p.
5. Kazanskii A.A. (2011). Object-oriented programming in Microsoft Visual C # on the platform of Microsoft Visual Studio 2008 and the .NET Framework. Part 3: Tutorial and Workshop [Ob'ectno-orientirovannoye programmirovanie na yazyke C#]. MSU of Civil Engineering. Moscow: 2011 (inRussian).
6. Online English-Russian Dictionary: http://dictionary.reverso.net/ english-russian.
7. Online dictionary: google translate: http://google.translate.com.
МОДЕЛИРОВАНИЕ СИСТЕМЫ МАШИННОГО ПЕРЕВОДА ТЕХНИЧЕСКИХ ТЕКСТОВ В ТЕЛЕКОММУНИКАЦИОННОЙ СФЕРЕ С ИСПОЛЬЗОВАНИЕМ ОБЪЕКТНО-ОРИЕНТИРОВАННОГО ЯЗЫКА ПРОГРАММИРОВАНИЯ С#
Алешинцев Андрей Владимирович, Московский технический университет связи и информатики, Москва, Россия,
Бессонова Елена Владимировна, Московский государственный строительный университет, Москва, Россия,
Сак Александр Николаевич, Московский государственный строительный университет, Москва, Россия, [email protected]
Дннотация
Необходимость в постоянном доступе к научным статьям или инструкциям на английском языке по применению телекоммуникационной техники и устройств, а также появление новых технологий нуждается в разработке системы машинного перевода для телекоммуникационной отрасли. Развитие языков объектно-ориентированного программирования позволяет представлять объекты, содержащие большое количество информации, наиболее удобным способом с точки зрения стоящих перед разработчиком задач. В статье рассматриваются вопросы представления лексических единиц предложения как объектов программы на языке С#, а также вопросы особенностей перевода соответствующих лексем, учитывая грамматику английского и русского языка в области телекоммуникаций. Авторы пытаются избежать представления лексико-семантической структуры предложения посредством метаязыка - посредника, значительно утяжеляющего систему перевода, путем создания более легкой и понятной структуры. Для этого устанавливаются определенные правила лексического и грамматического соотношения повествовательного предложения на английском и русском языках. Учитываются особенности обработки Ы:т1-страниц онлайн словарей, используемых как тезаурус лексических единиц при переводе.
Ключевые слова: машинный перевод, объектно-ориентированное программирование, лексема, грамматика, массивы, коллекции, онлайн словари.
Литература
1. Санников В.Г., Алешинцев А.В. Многочастотный модем как один из основных элементов системы "интеллектуальное здание" при удаленном управлении объектами // Т-Сомм: Телекоммуникации и транспорт. Т. 9. № 6. 2015. С. 21-27.
2. Сак А.Н. Идентификация членов предложения как основная задача машинного перевода // Научное обозрение. Вып. 14. 2015. С. 9-14.
3. Мельчук И.А. Язык: от смысла к тексту. М.: Языки славянской культуры, 2012. 176 с.
4. Мельчук И.А. Русский язык в модели "Смысл Ф Текст". Москва - Вена: Школа "Языки русской культуры", Венский славистический альманах, 1995. XXVIII. 682 с.
5. Казанский А.А. Объектно-ориентированное программирование на языке Microsoft Visual C# в среде разработки Microsoft Visual Studio 2008 и .NET Framework. Ч.3: Учебное пособие и практикум. М.: Моск. Гос. Строит. Ун-т., 2011. |80 с.
6. Онлайн англо-русский словарь: http://dictionary.reverso.net/english-russian.
7. Онлайн словарь Google translate: http://google.translate.com.
Информация об авторах:
Алешинцев Андрей Владимирович, старший преподаватель, Московский технический университет связи и информатики, Москва, Россия
Бессонова Елена Владимировна, к. филол. н, доцент, Московский государственный строительный университет, Москва, Россия Сак Александр Николаевич, к. филол. н, доцент, Московский государственный строительный университет, Москва, Россия
f I л
