Атты І халықаралық конференция ЕҢбектері
жүктеу/скачать 8,57 Mb. Pdf көрінісі
|
- Бұл бет үшін навигация:
- Those children are mine –O`sha bolalar meniki
| 1
2 3 4 5 I go to the university 1 5 3 2 Men universitet -ga - boraman Substitution proceeded through in two ways: 1) concretization; 2) generalization. These are proceeded in morphologic and synacticacts. For example, “and”– (va, hamda in Uzbek) is as connectable conjunction used the following functions: bo`lsa, esa, biroq – I shall go and you stay here –Men ketaman, sen bo`lsang shu yerda qol. It comes as infinitive (harakat nomi) in Uzbek in the compound units of verbs, as well: try and do it – buni qilishga harakat qiling, come and see –ko`rishga keling, wait and see –yashasak ko`ramiz [4]. Generalization may often occur mainly among the syntactic units. In replenishment process some morphologic categories are added by the position of speech. For instance, I have a book –Menda kitob bor. The affix –da is added to personal pronoun. It seems very easy if it is done by human. But for the linguistic database of translation it should be clarified accurately. In the process of omission one or several units are deleted in the context: demonstrative pronoun those –ular, o`shalar, ana o`shalar; these –bular, shular are used for plural noun, but sometimes as singular in Uzbek. Those children are mine –O`sha bolalar meniki. In English we may see the agreement of those and children (plural), but in Uzbek it is not normal. That’s why affix –lar (o`shalar) is omitted in this situation. There is some evidence to suggest that machine translation system has contextual and strong grammatical database. It is important to input morphological classification with equivalencies in two languages (English-Uzbek). We remind that Krosslexicon has morphological classificatory which holds 115 groups of declinable words in electron dictionary [5]. The morphoclassificatory can make the word forms even if they do not exist in the dictionary. We can obtain good results in this field in case grammatical peculiarities of the text are considered true. There are many problems in morphological level in bilingual program. Especially, it should be done formalization and modeling of linguistic database in Uzbek. Translation program consists of three stages: languageprocessor which contains analyser and synthesizer; linguistic model which contains of the knowledge of grammar and semantics; associative procedure which expresses linguistic translation operation that is connective between declarative and procedural parts [6] . We may observe that the module as analyze->transfer- >synthesis is proceeded in many machine translation systems. Analizator should be provided linguistically in the process of morphological analyze. Generally speaking, grammatical base of translation program has done in scientific researches. According to them there are following types of the grammar: 1) chain grammar (цепочечная грамматика); 2) component grammar (грамматика составляющих); 3) dependency grammar (грамматика зависимостей); 4) context- free grammar (контекстно-свободных грамматикa); 5) lexical-functional grammar (лексико- функциональная грамматика); 6) unificational grammar (унификационные грамматикa). “Chain grammar consists of words that belong to the groups of the terms (article+noun+preposition) and compounding units such as (subject+predicate) functional elements of terms. The order of units of speeches are shown in this. Component grammar providesthe group of grammar elements, for example, group of noun phrase (noun, article, adjective and other modifications), prepositional group (preposition +part of speech). In dependency grammar each elements are dependable each others. The strategy of analyze is as top-downand the center of sentence is predicate (verb). A transformational method is used in contextual grammar and it has shown above. Unification grammar consist of four components: suite of unification, interpreter for grammar rules and 302 description of the words, directed graphs of possessing program and analyzer with helping graph- devices. Unification grammar identifies semantic valency with synactic valency and description of dictionary with grammar rules [7]”. It is truly estimated by V.Rojdestvenskiy that central problem of artificial intelligence is machine translation [8].Because facilities of languageare are appeared by the influence of linguistic and nonlinguistic factors. Machine translation is complex physiological process. That’s why by using contemporary methods of the schoolars, we have to build well-built linguistic database of translation program. Sum up, powerful linguistic and programming database characterize the quality of machine translation. All grammar rules of the text must be investigated depending on types of style of the texts. It will be better if contextual dictionaries are created in English-Uzbek translation program. жүктеу/скачать 8,57 Mb. Достарыңызбен бөлісу:
|