Year: 2025
Type: dissertation
Abstract: The paper addresses the task of classifying humorous texts in Russian using the RuBERT model with fine-tuning, aiming to develop an effective system for identifying comic content. To achieve this goal, datasets of jokes and anti-jokes were created, data preprocessing was carried out, and the model was trained. Methods of deep learning, tokenization, and fine-tuning were employed. As a result, high performance metrics—accuracy and F1-score—were achieved, confirming the effectiveness of the approach. The model demonstrates potential for application in content moderation and user text analysis.
Authors: Лощинин, Д. С., Loshchinin, D. S.
Language: Russian
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7117193677