Tuplam (Xalqaro konferensiya-2022-BuxDU). pdf
Yüklə 1,07 Mb. Pdf görüntüsü
|
- Bu səhifədəki naviqasiya:
- Some of the authors of this publication are also working on these related projects
|
See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/363298655 Matnlarni intellektual taxlil qilish masalalari Conference Paper · September 2022 CITATIONS 0 READS 119 1 author: Some of the authors of this publication are also working on these related projects: data mining View project data base View project Batirbek Samandarov Karakalpak State University 11 PUBLICATIONS 15 CITATIONS SEE PROFILE All content following this page was uploaded by Batirbek Samandarov on 06 September 2022. The user has requested enhancement of the downloaded file. ЎЗБЕКИСТОН РЕСПУБЛИКАСИ ОЛИЙ ВА ЎРТА МАХСУС ТАЪЛИМ ВАЗИРЛИГИ ЎЗБЕКИСТОН РЕСПУБЛИКАСИ ФАНЛАР АКАДЕМИЯСИ В.И. РОМАНОВСКИЙ НОМИДАГИ МАТЕМАТИКА ИНСТИТУТИ ЎЗБЕКИСТОН МИЛЛИЙ УНИВЕРСИТЕТИ ТОШКЕНТ ДАВЛАТ ТРАНСПОРТ УНИВЕРСИТЕТИ БУХОРО ДАВЛАТ УНИВЕРСИТЕТИ Бухоро фарзанди, Беруний номидаги Давлат мукофоти лауреати, кўплаб ёш изланувчиларнинг ўз йўлини топиб олишида раҳнамолик қилган етук олим, физика-математика фанлари доктори Ғайбулла Назруллаевич Салиховнинг 90 йиллик юбилейларига бағишланади АМАЛИЙ МАТЕМАТИКА ВА АХБОРОТ ТЕХНОЛОГИЯЛАРИНИНГ ЗАМОНАВИЙ МУАММОЛАРИ ХАЛҚАРО ИЛМИЙ-АМАЛИЙ АНЖУМАН МАТЕРИАЛЛАРИ 2022 йил, 11-12 май БУХОРО – 2022 ТАШКИЛИЙ ҚЎМИТА Фахрий раислар: Аюпов Шавкат В.И.Романовский номидаги Математика Институти директори, академик Маджидов Иномжон М.Улуғбек номидаги Ўзбекистон Миллий Университети ректори Абдурахманов Одил Тошкент давлат транспорт университети ректори Хамидов Обиджон Бухоро давлат университети ректори Раислар: Розиқов Ўткир ЎзФА Математика Институти илм-фан бўйича директор ўринбосари, профессор Арипов Мирсаид ЎзМУ, профессор Шадиметов Холматвай Тошкент давлат транспорт университети, профессор Дурдиев Дурдимурод ЎзФА Математика Институти Бухоро бўлими мудири, профессор Раис ўринбосарлари: Ҳаётов Абдулло В.И.Романовский номидаги Математика Институти, профессор Худойберганов Мирзоали ЎзМУ, ф.-м.ф.д. Эшанқулов Ҳамза БухДУ, факультет декани, т.ф.ф.д. (PhD) ТАШКИЛИЙ ҚЎМИТА АЪЗОЛАРИ Жўраев А.Т. БухДУ, проректор Жумаев Р.Ғ. БухДУ, проректор Зарипов Г.Т. БухДУ, доцент Жумаев Ж. БухДУ, доцент Расулов Т.Ҳ. БухДУ, профессор Жалолов О.И. БухДУ, кафедра мудири, доцент Шафиев Т.Р. БухДУ, кафедра мудири, т.ф.ф.д.(PhD) Бабаев С.С. БухДУ, ф.-м.ф.ф.д.(PhD) Ахмедов Д.М В.И.Романовский номидаги Математика институти, (PhD) Болтаев А.Қ В.И.Романовский номидаги Математика институти, (PhD) Дурдиев У.Д. БухДУ, доцент Дилмуродов Э.Б. БухДУ, доцент Жумаев Ж.Ж. ЎзФА Математика Институти Бухоро бўлинмаси, (PhD) Зарипова Г.К. БухДУ, доцент Сайидова Н.С. БухДУ, доцент Бакаев И.И. Рақамли технологиялар ва сунъий интеллектни ривожлантириш илмий-тадқиқот институти, (PhD) Шадманов И.У. Математика Институти Бухоро бўлинмаси, (PhD) Хаятов Х.У. БухДУ, катта ўқитувчи Хазратов Ф.Х. БухДУ, катта ўқитувчи Эргашев А.А. БухДУ, катта ўқитувчи Авезов А.А БухДУ, катта ўқитувчи 449 the control of the reliability of customs declarations is in the field of economic interests of any state, due to the fact that customs payments, as a rule, make up a considerable part of the revenue side of the state budget. For example, over the past five years, customs payments amounted to 13-18% of the total revenue part of the state budget of the Republic of Uzbekistan [1]. Therefore, the task of investigating methods for identifying unreliable customs declarations is relevant. Information reliability control methods The task of identifying unreliable customs declarations is a special case of the general and, as you know, ancient task of identifying false information, i.e. how to distinguish "truth" from "falsehood". Numerous scholars have been interested in this problem, starting with the ancient Greek philosopher Aristotle (384 B.C.). Also famous are Leibniz's "sufficient reason" Law, Godel's Incompleteness Theorem, Kolmogorov's Superposition Theorem, Satoshi Nakamoto's Blockchain System Technology, Xin Luna Dong's "Knowledge-based Reliability" Theory, and others. Not only scientists of natural or secular sciences were interested in this problem. This refers to the scholars who were involved in determining the authenticity of hadith in Islam. The legendary Muhammad ibn Ismoil al-Bukhari (IX century) is considered to be one of them. He is said to have devoted his entire life to collecting and analyzing hadith, and he worked to develop methods that allowed him to identify "authentic" hadith. Having analyzed over 600,000 hadiths, he chose only 7,275 (1.21%) for his book, Al- Jami as-Sahih, which has been considered the most reliable book for over 11 centuries. According to the conditions of Imam Bukhari, the criteria for the authenticity of hadith consist of two groups of conditions: - Conditions for the source of the hadith; - conditions for the content of the hadith. Based on the concept of Imam Bukhoriy, the authors of this paper have developed criteria for the reliability of customs information, consisting of two groups and six conditions: - conditions for the source of customs information; - conditions for the content of customs information. Conclusion. In conclusion, we would like to note that the above-mentioned criteria for ensuring the reliability of customs information are implemented in the automated information systems of the State Customs Committee of the Republic of Uzbekistan and have shown their effectiveness. As a result of the functioning of this system during 2020 revealed 201 thousand 611 unreliable cargo customs declarations and charged to the state budget additional 36.7 billion UZS customs payments. REFERENCES 1. Saidov A.A. Classical methods of controlling the reliability of information and peculiarities of their application to the customs case // Monograph. -Tashkent. -2021. – 498 p. https://elibrary.ru/item.asp?id=47759425. MATNLARNI INTELLEKTUAL TAXLIL QILISH MASALALARI Samandarov B.S., To‘xtabaev U.A., Ispanova J.P. Berdaq nomidagi Qoraqalpoq davlat universiteti, O’zbekiston Keyingi yillarda lingvistik tahlil va matnni qayta ishlash sohasida ko‘plagan tadqiqotlar olib borilib [1], bunda statistik usullar yordamida mualli fning uslubini o‘rganish yoki axborot tizimlarida yangiliklar yoki blog materiallari reytingini o‘rganish, foydalanuvchi xatti-harakatlarini modellashtirish tadqiq qilib kelinmoqda [2]. Bugungi kunda Internet va boshqa manbalarda axborotlar keskin o‘sishda davom etmoqda va bu axborotlarning asosiy qismi matn ko‘rinishida shakllantirilmoqda. Ushbu katta xajmdagi ma’lumotlardan foydalisini ajratib olish uchun maxsus usul va mexanizlar zaruriyati kelib chiqadi [3]. Matndan belgilarni ajratib olish (Feature Extraction) – Data Mining sohasining keng uchrashadigan masalasi bo‘lib, aynan belgilarni generatsiya qilish bosqichida qo‘llaniladi. Shundan kelib chiqib, bugungi kunda matnlarni intellektual tahlil qilish Text Mining nomini oldi. Bu holatda Feature Extraction – NLP (NLP, Natural Language Processing – tabiiy tilga ishlov berish) sohosiga tegishli bo‘lib, sun’iy intellekt va matematik lingvistikaning alohida sohasi sanaladi. Bugungi kunda NLP ko‘lami nutqni aniqlash va matnni tarjima qilishdan bosh lanib, bashoratli matn kiritish va mashina va odam o‘rtasida aloqa o‘rnatishgacha qo‘llaniladi. Bunda mashinani o‘qitish (Machine Learning) yordamida matn ma’lumotlarini tanib olish va tahlil qilish vazifalari hal qilinadi hamda matndan belgilarni ajratib olish va data scientist 450 mutaxassis ma’lumotlarni tayyorlashning (Data Preparation) ushbu bosqichini qanday amalga oshirishi tahlil qilinadi. Matndan belgilarni ajratib olishda dastlab uni ML (Machine Learning) algoritmlari yordamida ishlov berishga yaroqli ko‘rinishga keltirish zarurati kelib chiqadi. Buni amalga oshirish uchun matnlarni bir nechta bosqichda qayta ishlash zarurati kelib chiqadi: ▪ Tokenlash – matnning uzun qismlarini kichikroq qismlarga bo‘lish (paragraflar, jumlalar, so‘zlar). TextMingda tokenlash matnni qayta ishlashning birinchi sanaladi. ▪ Normallashtirish – matnni «tozalangan» shaklga keltirish, ya’ni, so‘zlarning yagona reestri, tinish b elgilarining yo‘qligi, shifrlangan qisqartmalar, raqamlarning og‘zaki yozilishi va boshqalar; ▪ Stemlash - qo‘shimchalarni (qo‘shimcha, prefiks, tugatish) olib tashlash orqali so‘zni o‘z ildiziga olib kelish; ▪ Lemmalash - so‘zni semantik kanonik shakliga qisqartirish; ▪ Tozalash - semantik yukni ko‘tarmaydigan to‘xtash so‘zlarini (so‘z birikmalari, qo‘shma gaplar va boshqalar) olib tashlash. Ushbu amallar orqali matnlar zarur belgilarni ajratib olish uchun sonli qayta ishlashga tayyor ko‘rinishga keltiriladi. B izga sir emaski, ko‘plagan kompaniyalarda juda katta miqdorda foydalanilmaydigan ma’lumotlar (dark data ) mavjud. IBM kompaniyasi hisob- kitoblariga ko‘ra datchiklar va analog raqamli konvertorlar tomonidan shakllantirilayotgan ma’lumotlarning taxminan 90% xech qachon ishlatilmaydi. Xulosa qilib ayttadigan bo‘lsak, strukturalanmagan foydalanilmaydigan ma’lumotlardan foydali ma’lumotlarni ajratib olib ularga ishlov berish eng istiqbolli usullardan biri bo‘lib qoladi. ADABIYOTLAR 1. В.Э.Пашковский, В.Р.Пиотровская, Р.Г.Пиотровский. Психиатрическая лингвистива Изд. 4- е. – М.: ЛЕНАНД, 2015. – 168 с. 2. Gir´on J., Ginebra J., Riba A. Bayesian Analysis of a Multinomial Sequence and Homogeneity of Literary Style // The American Statistician. 2005. Vol. 59. Issue 1. PP. 19 –30 3. Faiyaz Ahmad, Yassar, Amreen Ahmad. Automatic Summarization of Textual Document // International Journal of Innovative Technology and Exploring Engineering (IJITEE), Volume-9 Issue-1, November 2019. PP 2486-2491 PREDICTING AND CLASSIFYING OF PUPILS' KNOWLEDGE USING MACHINE LEARNING ALGORITHMS Samandarov E.K. National university of Uzbekistan, Tashkent, Uzbekistan Educational Data Mining is a rising discipline that aims to develop methods to explore data from educational contexts, i n order to understand students’ behavior, interests and results in a better way. The EDM can be considered as the intersection of three main areas: education, statistics and informatics. This intersection among these three areas also generates other sub-fields, narrowly related to EDM, such as computer-based education, learning analysis, data mining (DM) and machine learning (ML). As seen in Fig. 1, machine learning is critical in educational data mining. It gives the capacity to forecast in the educational sector. One advantage of this approach is that it can identify recurring queries. There are four main types of machine learning: supervised, unsupervised, semi-supervised and reinforcement. The supervised learning is the most popular paradigm for machine learning, it aims to build a model based on observation data and desired results; This model allows the best approach to the relationship between input and output observable in the data. Supervised learning problems can be grouped into regression and classification problems. While, unsupervised learning uses only input observations data without having corresponding outputs. So, the goal of unsupervised learning is to determine the patterns or clusters hidden in the data from unlabeled data. Unsupervised learning problems can be seen as a problem of clustering or association. The semi-supervised learning falls between the two previous types of learning. Indeed, data labeling is a very expensive operation and requires presence of human experts. However, the availability of labels in some observations, even if they are missing in the most cases, gives to the semi-supervised algorithms the best chance for model construction. The semi- supervised learning algorithms exploit the idea that, even if group memberships of unlabeled data are Dusmukhametov A.I., Saidov A.A., Khakimova F.A.PROBLEMATIC ISSUES OF CUSTOMS CONTROL ORGANIZATION RELATED TO THE USE OF ARTIFICIAL INTELLIGENCE METHODS ............................................................................................................................................. 437 Ergashev A.А., Kayumova N.N. MA`LUMOTLAR BAZASINING TAHLILIY IMKONIYATINI OSHIRISH .............................................................................................................................................. 438 Eshankulov H.I., Sultonov H. TAQSIMLANGAN AXBOROT TIZIMLARNING ARXITEKTURASI ......................................................................................................................439 Eshonqulov H.I. IDEOLOGY OF ONTOLOGY WEB LANGUAGE .................................................... 440 Ibragimov S. MODELING INDIVIDUAL LIFE TRAJECTORIES BY GRAPH .................................. 441 Ibragimov Sh.M. ARTIFICIAL INTELLIGENCE – DEVELOPMENT PROSPECTS..........................443 Ismoilova D. ASSOTSATSIYA QOIDASINI O’RGANISH VA QO’LLASH.......................................443 Polvonov S.Z., Akramov O. I. PYTHONDA LOGISTIK REGRESSIYA ALGORITMINI AMALGA OSHIRISH .............................................................................................................................................. 444 Qobilov K.H., Olimov N.N., Toyirova U.I. SUN’IY INTELLEKT MASALALARINI YECHISH MODELLARI ......................................................................................................................................... 446 Risqaliyev J.D. SUN’IY INTELLEKTDA MANTIQIY REGRESSIYANING O‘RNI ..........................447 Ro’zimatov S. Sh., Rahimov A. G’. TA'LIM TIZIMIDAGI SUN'IY AQLNING KELAJAGI ..............448 Saidov A.A., Khakimova F.A., Abdurakhmanov T.T. APPLICATION OF THE CONDITIONS OF IMAMA BUKHARIY TO MODERN INFORMATION CHALLENGES ............................................. 448 Samandarov B.S., To‘xtabaev U.A., Ispanova J.P. MATNLARNI INTELLEKTUAL TAXLIL QILISH MASALALARI ...................................................................................................................................... 449 Samandarov E.K. PREDICTING AND CLASSIFYING OF PUPILS' KNOWLEDGE USING MACHINE LEARNING ALGORITHMS .............................................................................................. 450 Xazratov F. X., Rufatov J. Z., Boltayev S.B. BIG DATA VA MA`LUMOTLAR TAHLILI TURLI SOHALARDA QO`LLANILISHI........................................................................................................... 451 Бакаев И.И., Бакаева Р.И. СОЗДАНИЕ АЛГОРИТМА ТОКЕНИЗАЦИИ НА ОСНОВЕ БАЗА ЗНАНИЙ ДЛЯ УЗБЕКСКОГО ЯЗЫКА. ............................................................................................. 452 Болтаев Т.Б., Ибрагимов С.И. СПЕЦИФИКАЦИЯ ИНТЕЛЛЕКТУАЛЬНОГО АНАЛИЗА ПРОЦЕССА ОБУЧЕНИЯ НА ОСНОВЕ МОДЕЛИ ДИДАКТИКИ.................................................. 453 Гаращенко А.В., Эргашев Н.Х . ФОРМИРОВАНИЕ ИНДИВИДУАЛЬНОЙ ОБРАЗОВАТЕЛЬНОЙ ТРАЕКТОРИИ НА ОСНОВЕ МНОГОУРОВНЕВОЙ CNN-LSTM СИСТЕМЫ НЕЙРОННЫХ СЕТЕЙ ........................................................................................................................................455 Ғайбулов Қ.М. ҚАРOРНИ ҚЎЛЛАБ-ҚУВВАТЛАШ ТИЗИМЛАРИНИ (ҚҚҚТ) ҚУРИЛИШ МАТЕРИАЛЛАРНИ ТАНЛАШГА ҚЎЛЛАШ ................................................................................... 456 Кодиров З., Студенкова Д., Косимов Д. ИСКУССТВЕННЫЙ ИНТЕЛЛЕКТ В ПОМОЩЬ ПРЕПОДАВАТЕЛЮ............................................................................................................................. 457 Сеитназаров К.К., Туремуратова Б.К. ОСОБЕННОСТИ ИЗУЧЕНИЯ ИНТЕЛЛЕКТУАЛЬНЫХ АГЕНТОВ ДЛЯ ПРОГРАММИРОВАНИЯ ........................................................................................ 457 Туремуратова Б.К., Кенесбаева Д.А. ОСНОВНЫЕ НАПРАВЛЕНИЯ РАБОТ В ОБЛАСТИ ИСКУССТВЕННОГО ИНТЕЛЛЕКТА................................................................................................ 458 Эргашев А. A., Холиков А.О. МИЖОЗ СEРВEР ТEХНОЛОГИЯЛАРИДА ИЛОВАЛАРНИ ИШЛАТИШ УЧУН MICROSOFT AZURE АСОСИДАГИ БУЛУТ ТEХНОЛОГИЯЛАРИДАН ФОЙДАЛАНИШ...................................................................................................................................459 VII ШЎЪБА. АХБОРОТ ХАВФСИЗЛИГИ. INFORMATION SECURITY.................... 460 Adizova Z.M., Davletov J. K. PYTHON DASTURLASH TILI ORQALI AXBOROT XAVSIZLIGINI TAMINLASH. ........................................................................................................................................ 460 Eshonqulov Sh. XODIMLARNI FACE ID YORDAMIDA BIOMETRIK AVTORIZATSIYADAN O‘TKAZISH AXBOROT TIZIMINI TASHKIL ETISHNING TEXNIK TALABLARI ....................... 460 Matyakubov A.S., Tadjiev R.N., Komilov R.K. KIRUVCHI VA CHIQUVCHI TARMOQ TRAFIGINI TEKSHIRISH VA BOSHQARISHNING ILG’OR USULLARI............................................................ 461 Mavlonov Sh. H., Baxramov M. S. KIBERJINOYATCHILIKKA QARSHI KIBERXAVFSIZLIK ..... 462 Mavlonov Sh.H. ZAMONAVIY RAQAMLI TEXNOLOGIYALARDAN FOYDALANISHDA KIBERJINOYATCHILIKNING OLDINI OLISH.................................................................................. 463 Mirzakulov J. POSTGRESQL - DATABASE FOR HIGH PROTECTION. .......................................... 464 Nurullayev M.M. KRIPTOGRAFIK KALITLARNI SHAKLLANTIRISH UCHUN TASODIFIY SONLARNI GENERATSIYALASHDA SMARTFON SENSORLARIDAN FOYDALANISH ........... 465 View publication stats Yüklə 1,07 Mb. Dostları ilə paylaş:
|