Journal of Word Art – Сўз санъати халқаро журнали – Международный журнал искусства слова 6-том 1-нөмір (2023)
NLPNING ZAMONAVIY ALGORITMLARI VA KONSEPSIYALARI
Элов Ботир Болтаевич, Хамроева Шахло Миржоновна, Хусаинова Зилола Юлдашевна
Аңдатпа
Natural language processing (NLP) algorithms serve to process human language data, including unstructured text data. Today, NLP algorithms are developed based on language rule-based, statistical and artificial intelligence approaches. Based on the approach based on language rules, the formation of linguistic bases for NLP tasks and the operations of classification in language corpora are performed. Statistical algorithms allow machines to read, understand and derive meaning from human languages and are based on processing large volumes of (bigdata) texts. Statistical algorithms are used in many NLP tasks such as speech recognition, machine translation, sentiment analysis, text classification and analysis. Today, deep learning models of machine learning (ML) algorithms based on CNN and RNN technologies allow to "learn" existing NLP systems and allow more accurate processing of large volumes of unstructured texts. This article discusses modern algorithms and concepts of NLP today, and methods of processing texts in the Uzbek language based on these algorithms are given.
NLP, конвейерный конвейер, расстояние Левенштейна, косинусное сходство, метод мешка слов, алгоритм TF-IDFNLP, pipeline pipeline, Levenshtein distance, cosine similarity, Bag of words method, TF-IDF algorithm.NLP, pipeline konveyeri, Levenshtein masofasi, kosinus o‘xshashligi, Bag of words usuli, TF-IDF algoritmi.
Метадеректер дереккөзі: журналдың OAI-PMH архиві · Sindex толық мәтінді сақтамайды, дереккөзге сілтеме береді.