The growing popularity of Chinese social media platforms such as Sina Weibo has created a large number of user generated text content, which is of great value for understanding public emotions. However, the existence of mixed languages in these texts, especially Chinese and English, and mixed expressions pose a major challenge to current emotion classification methods. To address these issues, we propose a Bilingual Feature Fusion Network (BFFN) that leverages the multilingual capabilities of pre-trained language models to enhance the semantic feature extraction of Chinese text. Additionally, we introduce a Bilingual Cross Attention Mechanism (BCAM) that utilizes emotional features as the primary factor to capture cross-lingual emotional information effectively. Furthermore, we employ a lightweight fine-tuning approach that combines Low-Rank Adaptation (LoRA) and Embedding Fine-tuning (LEF) to reduce the complexity of fine-tuning model weights for downstream tasks. Extensive experiments on various datasets demonstrate the superiority of our proposed method, outperforming state-of-the-art models like ERNIE by 1.43% in accuracy. Our work contributes to the advancement of emotion classification in the context of mixed-language communication culture and provides a practical solution for real-world applications. Our code has been published on the open source community Github ( https://github.com/oujieww/BFFN ).