N.Z. ABDURAKHMONOVA
National University of Uzbekistan named after Mirzo, Tashkent, Uzbekistan
GRAMMATICAL ANALYZE IN MACHINE TRANSLATION
BETWEEN ENGLISH AND UZBEK
Today human society characterizes high degree of activity in different fields such as economy,
science, technology, culture etc. And it has caused to increase body of information that presents
some difficulties between person to person or among the group of people. Computer has being
considered one of the main approaches to ease opportunities of people since it was invented. So
machine translation is used to exchange information communicative attitudes. Translation of the
text is very complex creative process from a natural language into another one. We can see now
variety forms of machine translation system; even they can recognize speech and translate orally.
Most of them are multilingual translation programs. The Uzbek language is being developed rapidly
300
after our independence. Therefore our research has taken the first step to build of linguistic database
of translation program.
It is always noticed to mainly grammatical analyze (morphological and syntax) of the lexemes in
any system (retrieval of database, machine translation, automatic editor). So analyze is investigated
as one of the base of linguistic approaches.
There are many methods such as tokenization, lemmetization, stemming which used in the
automatic morphologic analysing process for the texts. Tokenization helps to detach the meaning
of units of speech (token, wordform) separately.The previous forms of the words are determined in
lemmatization process. One of the processes is stemming. The roots of the words are found by its
assist. Three ways which are used for analyze, we may describe like a chart:
In some literatures above mentioned methods are used as terms which are special procedures and
programs of creating corpus – database of computer software
[1]
.
Our decent research has shown the necessity of morphoclassificator in the automatic process to
translate from English texts into Uzbek. Morphological classification of the words might be taken
as the main way to clarify part of speech. We admit that two languages belong to other language
family.Some issues are demanded that to solve to input the linguistic database to computer
software. We cite an example contrasting between adjective and adverb in English with equivalency
in the Uzbek language: It is a good impression –Bu ijobiy taassurot. He speaks English well (badly)
–U ingliz tilida yaxshi (yomon) gapiradi. Both of the words (yaxshi, yomon) are considered as
adjective. But they are only analyzed as modifier in syntax not in morphology as adverb. Focus on
the problem is these words are not existed in adverbial list in the process. As well as we can face to
again other examples between adjective and noun: I like to eat wooden bowl –Men yog`och kosani
yoqtiraman. In this place “yog`och”is not “wood” but “wooden”.
Naturally, English and Uzbek are member of different type of language so their linguistic nature
is diverse too. For instance some pronounces in English don’t exist in Uzbek, and they are called as
other categories: few, little words are used as adverb such as “kam, oz” in Uzbek. Such problems
seem easy to solve at first. It depends on not only electron dictionary which are decoded in the
languages, but also it is responsive to grammatical analyze in context. One is of the urgent request
for any translation to save agreement between the form and sense of the text.
Transformation method is estimated as effective way to solve the problems. Four stages proceed
in the transformation method: transposition, substitution, replenishment, and omission [2]. But other
author presents only three ways: adjunction, substition, deletion. In addition to this, the base of the
transformation process contains kernel structures, and it consists of simple sentences in syntax [3].
Grammatical analyze
Tokenization: token–children; token–more
Lemmatization: lemma –child; lemma –many| much
Stemming: stem –child; stem –many| much
301
These main stages signifies in the process of analyze and synthesis. We analyze the bases on the
examples abovementioned types of transformation method.
In the process of transposition it has been analyzed words and word combinations in the text.
Достарыңызбен бөлісу: |