The evolution of Chinese lexical meanings in traditional cultural texts is closely related to the cultural and epochal background of the language. In this paper, a corpus of Chinese word meanings in traditional cultural texts is constructed by combining Wang Li’s Dictionary of Ancient Chinese and Dazidian of Chinese. A two-way LSTM language model and ELMo vector generation algorithm are used to represent word vectors, and a dynamic vocabulary representation learning model is constructed to realize the task of polysemous word recognition. And the cosine similarity is used to calculate the similarity between the sets of related words before and after a certain time node, and then to determine whether the lexical meaning of a word has changed. When the obtained similarity value is larger, it indicates that the change of word meaning is small or does not occur, and vice versa, the change of word meaning is large. The results of the study show that the average accuracy of the Lexical Representation Learning Model (ELMo) is around 75%. Taking the character “Shai” as an example, it can be seen that the character “Shai” is closer to its original meaning from the pre-Qin to Han dynasties, and evolved the meaning of cultural display from the Wei, Jin, Southern and Northern Dynasties to the Tang Dynasty. In the Song, Yuan, Ming and Qing Dynasties, the character “Shai” began to become vernacular. Changes in time and social development are important factors contributing to the evolution of Chinese word meanings in traditional cultural texts.
