Lingvist oxirgi shakl 1-son indd
Download 1.51 Mb. Pdf ko'rish
|
Lingvist OXIRGI SHAKL 1-son
- Bu sahifa navigatsiya:
- For citation
Keywords: artificial intelligence, ChatGPT, GPT-4, Dragonfly, Claude,
Claude-instant, NeevaAI, artificial language, corpus, parallel corpus, corpus types, multimedia-corpus, automatic translation, compatibility, a database. Computational linguistics is developing in world linguistics. Many of the tasks set for computational linguistics at the end of the 20th century are being solved today. One of them is the problem of artificial intelligence. Artificial intelligence is one of the high-level products of the information age and is used in all areas of production. Today’s workers, household robots, the foundations of nanotechnology are a manifestation of the capabilities of artificial intelligence. Artificial intelligence is a product of human thinking. With the help of machine memory, artificial intelligence performs not only physical movements, but also processes associated with human mental abilities. «Artificial intelligence is the ability of a computer or robot 1 Aziza Rahkmanova – For citation: Rahkmanova, A. 2023. “The role of parallel corpora in the improve- ment of artificial intelligence”. Uzbekistan: Language and Culture. Linguistics. 2023/1 (1): 153-162. Uzbekistan: Language and Culture 2023/1 (1) 153 to solve problems related to human mental activity, in particular, thinking, understanding the content of speech, summarizing information, etc. This term is also used to refer to a branch of computer technology related to the development of systems with these characteristics». [www.wikipedia]. Today, about twenty forms of artificial intelligence are used. ChatGPT, GPT-4, Dragonfly, Claude, Claude-instant, and NeevaAI artificial intelligence are referred to as the most active artificial intelligence. Portraits, landscape samples, illustrations are being created through artificial intelligence. Artificial intelligence has the ability to embody both real reality and unreal events. Programs that work with the help of artificial intelligence are called artificial languages. «Artificial language is a language formed in the fields of mathematics, informatics and linguistics based on the processing and formalization of natural language, which ensures the relationship between a computer and a person, is entered directly into the computer’s memory, and is used to implement the functions of the computer in the linguistic sphere» [Kholmanova, 2019:12]. For example, ChatGPT (Generative Pre-trained Transformer) is designed to create a system of responses to human requests based on language models. This artificial language allows you to communi- cate and talk with people. All AI visualizations, especially ChatGPT, are based on a large database of English-language internet texts. As a result of regular repetition of the processes carried out by artificial intelligence, the processed natural language features are mastered. Artificial intelligence makes it possible to deeply understand human knowledge and culture. As an artificial intelligence model, ChatGPT can answer questions, provide information on any topic, and has the ability to communicate. He can create text and compose poetic verses. Artificial intelligence models have common characteristics in understanding appeals and accepting requests. But they differ from each other in some features. Each model has its own characteristics in terms of the use of words and their placement in a certain order. That is why artificial intelligence differs according to the way and form of answering questions. Artificial intelligence is able to recog- nize each other based on the methods of writing and answering. “Artificial intelligence is created by humans. Therefore, it has its limitations. For example, artificial intelligence such as Sage, GPT- 4, Claude, Claude-instant, ChatGPT, NeevaAI, Dragonfly mentioned above provide detailed, complete answers to questions in English. Other languages show relatively limited operations. This situation Aziza RAKHMANOVA 154 was determined by the language in which the artificial intelligence learned the most information and in which language it performed the most logical operations. In order for artificial intelligence to correctly understand and interpret texts in the Uzbek language, many operations must be per- formed on the text in this language. The database in this language should be mastered. “Language based on a database is necessary for artificial intelligence, just as language is necessary for human com- munication and speech.” [Kholmanova, 2023:32]. The development and improvement of artificial intelligence is emphasized in the literature. Corpora are one of the factors for im- proving artificial intelligence, and one of the tools that create a wide range of opportunities for automatic translation. In recent years, a number of theoretical and practical achieve- ments have been made in corpus linguistics, which is a branch of Uzbek computational linguistics. Corpus is “a collection of texts in electronic format based on finding words, phrases, grammatical forms, word meanings through a specific search engine” [http://rusorpora.ru]. The corpus includes phonemes, graphemes, morphemes and larger units - lexemes, sen- tences and texts (artistic or scientific works, newspaper and maga- zine texts) [Baranov, 2001:61; Rahimov, 2011]. The corpora are, first of all, divided into a corpus of texts in the same language and a corpus of parallel texts [Rakhmonova, 2020; 32]. Download 1.51 Mb. Do'stlaringiz bilan baham: |
Ma'lumotlar bazasi mualliflik huquqi bilan himoyalangan ©fayllar.org 2024
ma'muriyatiga murojaat qiling
ma'muriyatiga murojaat qiling