This project is a Natural Language Processing (NLP) course design that explores the application of BERT and Chebyshev KAN layers in text sentiment analysis. The goal is to enhance sentiment analysis performance by integrating Chebyshev KAN layers into the BERT model. 本项目是一个自然语言处理（NLP）课程设计，探讨了 BERT 和 Chebyshev KAN 层在文本情感分析中的应用。目标是通过将 Chebyshev KAN 层