Recognition And Evolution Of Hot Topics In Hydrogen Production Field In China Based On Text Mining

At present China is gradually improving the system of hydrogen policies,and accelerating the development process of all links of China’s hydrogen industry.As the basis of the whole hydrogen energy industry chain and the premise of large-scale industrial development,hydrogen production has become an emerging field with great development potentials.It has potential economic growth points and infinite growth potential.Scientifically and efficiently identify the research hot spots in the field of hydrogen production and reveal the evolution process and change law of hot topics,which is conducive to supporting the discipline knowledge system in the field and is of great significance to promote the development and progress of the field.Therefore,based on the text mining method,this paper uses LDA topic model and Word2 vec tool to research the identification and evolution of hot topics in the field of hydrogen production in China.Firstly,this paper analyzes the influence characteristics and attention characteristics of hot topics,uses the output information of LDA theme model to obtain the annual number of articles published by each topic,so as to quantify the influence level and attention level of topics.The identification index system of feature dimension of hot topics in this paper makes up for the deficiency that the existing methods do not consider the characteristics of hot topics.Secondly,based on the hot topic identification of the whole corpus,this paper constructs the evolution analysis model,makes time slices and topic extraction of the text according to the temporal changes of the literature,and identifies the subtopics of hot topics in different time windows through similarity calculation.then it uses Word2 Vec to train keywords and main path vocabulary of the subtopics.This paper analyzes the structure evolution of hot topics by connecting subtopics in adjacent time windows,analyzes the content evolution of hot topics by connecting subtopic keywords in adjacent time windows on the main path,to provide an effective and feasible idea for the evolution analysis of hot topics in the field with insufficient text quantity.The identification index system of the feature dimension of hot topics and the evolution analysis model constructed in this paper are applied to the field of hydrogen production in China.This paper finds the hot topics with both influence and attention in the field through index calculation,and by comparing with co-word analysis,the effectiveness of the index system for hot topics identification is verified.At the same time,using the hot topic evolution analysis model,this paper analyzes the structure evolution and the content evolution in the field,realizes a more refined expression of topic evolution,and provides theoretical support for promoting the innovative development and application of hydrogen production in China.