Emotion analysis in low resource domain based on transfer learning and meta-learning methods aims to improve the ability of emotion analysis in low resource domain by using the performance of transfer learning and meta-learning. Among them, the extraction of task data in the initial domain of transfer learning and the transformation, methods and categories of task data in the new domain are the main contents of the application of transfer learning, and the “learning about learning” of meta-learning is the main process of learning. Then, the two learning methods are applied to the emotion analysis process in the low-resource domain as the basis. It can effectively enhance the ability of domain data extraction and conversion in low-resource fields, and at the same time strengthen the ability of emotion learning in low-resource fields, so that the analysis level can reach a certain height, so as to realize the development and application of sentiment analysis in low-resource fields.