Metaverse Topic Mining and Evolution Based on LDA Model
The metaverse literatures(retaining the abstract part of the literature)in the WoS core database were retrieved,the LDA(latent Dirichlet allocation)model was built based on Python.The optimal number of topics through perplexity and topic consistency are determined,and the 11 topics with the highest word frequency under the text were extracted.The topics related to metaverse were extracted and mined,and WordCloud technology and Sankey diagram(topic similarity)were used to analyze the topic evolution,so as to understand the development status and evolution trend of metaverse.The results show that the current research on metaverse is mainly in the aspects of"metaverse technology"and"metaverse application",and the keywords"technolog"and"virtual"are more prominent in the word cloud analysis.From the perspective of topic heat analysis,Topic 4(education metaverse research),Topic 7(metaverse technology research),and Topic 8(metaverse application research)are the hot spots of metaverse research.From the perspective of topic evolution analysis,in the first phase(from 2018 to 2022),topic 2 metaverse model research and topic 8 metaverse technology research showed strong evolutionary capabilities,while in the second phase(from 2023 to 2024),topic 4 metaverse experience research and topic 8 metaverse scene construction research showed strong evolutionary capabilities and high similarity with adjacent topics.
万方数据
维普
