Fundamentals of natural language processing by artificial intelligence technologies

The goal of the course “Fundamentals of Natural Language Processing with Artificial Intelligence Technologies” is to immerse bachelor’s students in a field that is relevant for future employment and further study. The course is designed for students of the humanities faculties at Lomonosov Moscow State University. Graduates of the course (spring 2023) took part in a meeting with representatives of developers of domestic GPT technologies for humanities students; as a result, several of them completed internships at Yandex in the summer of 2023. For example, on November 15, 2023, within the framework of the interfaculty course “Large Language Models in Humanitarian Research” (Avramenko A. P.), Olga Zagladila, a 2023 course graduate, shared her story of getting employed as an AI trainer in the summer of 2023. During the course, students will:

- gain knowledge of the principles of evaluation from the Turing Test to modern benchmarks; the foundations of the philosophical “Chinese Room” experiment; the principles of machine translation; the principles of psychometric analysis of user behavior (including in an interactive learning environment); techniques for tokenization, lemmatization, and text vectorization; the specifics of organizing and structuring Python code; the specific characteristics of the two main text processing tasks (text classification and named entity recognition, NER); and the architectural features of large language models based on pretrained transformers;

- learn to perform semantic analysis using electronic corpora; carry out basic text cleaning tasks (converting to lowercase or uppercase, spell‑checking, part‑of‑speech tagging); access the open‑source code of available language models for text processing; interpret the results of text processing by neural networks; and evaluate tools for machine‑assisted written and spoken translation;

- gain experience working with the following programs and websites: the National Corpus of the Russian Language; the programming environment provided by Yandex; open repositories of domestic developers on GitHub; machine translation algorithms; and tools such as Skell from SketchEngine, MyStem, GephiLite, etc.

The course program includes 16 sessions: 8 lectures and 8 seminar‑workshops, as well as 8 practical homework assignments

Requirements for students:

• knowledge of the basics of linguistics;
• knowledge of the basics of applying digital technologies in linguistics.

Telegram

Занятия проводятся на факультете иностранных языков и регионоведения МГУ им. М. В. Ломоносова

В программе курса 16 занятий: 8 лекций и 8 семинаров-практикумов, а также 8 практических домашних заданий

Формат проведения: офлайн

Старт курса: 13 февраля 2024 года

Занятия будут проходить по вторникам с 13.00 до 14.30

Telegram

Набор на курс 2023/2024 года закрыт