Research on Chinese word sense disambigution based on Prompt learning

Word sense disambiguation is an important research topic in natural language processing. Its purpose is to determine the word meaning of a polysemous word in different contexts. In order to improve the efficiency of disambiguation, a Chinese word sense disambiguation model PL-WSD based on prompt learning is proposed. The model uses the knowledge base HowNet to obtain the representation of each word meaning in polysemous words, uses the way of prompt learning to calculate the similarity score of each word meaning in a specific context, and obtains the word meaning with the highest score as the final result of polysemous words. Experiments show that PL-WSD model is greatly improved compared with other models.