Research on Tibetan Weibo Sentiment Analysis Based on Deep Learning

This paper focuses on the task of Tibetan Weibo sentiment analysis based on deep learning. It aims to conduct performance comparison experiments among three deep learning models: TextRNN, TextCNN, and TextRCNN, using a dataset comprising 10,000 Tibetan Weibo posts. The experimental results demonstrate that the TextRCNN model exhibits the best performance in Tibetan Weibo sentiment analysis, achieving an accuracy rate of 85%. This notable advantage is primarily attributed to the TextRCNN model's integration of RNN (Recurrent Neural Network) and CNN (Convolutional Neural Network) characteristics, enabling it to simultaneously capture both contextual information and local features of Tibetan text, thereby more accurately understanding and judging sentiment orientations in Weibo posts.