| In the 1960 s,the research focus of bibliometrics changed from traditional bibliographic analysis to citation analysis gradually due to the emergence of citation databases.Citation analysis focused on the citation links and citation rules between research papers.With the continuous development of Internet technology,some fulltext databases have begun to provide full-text data of structured papers in XML format,providing a data basis for full-text citation analysis.In addition,the continuous maturity of text analysis technologies such as natural language processing,text mining,and machine learning provides a guarantee for full-text citation analysis at the technical method level.The full-text citation analysis researches from the chapter structure,chart function and citation information of the paper.The research perspective of citation information analysis includes citation location,citation motivation and citation sentiment.Citation sentiment analysis refers to excavating citations from citation content.The authors' attitudes and emotional tendencies towards the cited literature have become a research focus in the field of citation information analysis.Citation sentiment analysis has theoretical significance.Besides,citation sentiment analysis is widely used in many aspects of scientific research in the practical level,which helps to promote the development of bibliometrics,deepen the function of citation analysis,and improve the understanding of academic paper writing.Citation sentiment analysis is a branch of the field of citation information analysis,which expands the research content of citation information analysis,provides new ideas for full-text citation analysis,and helps promote the development of bibliometrics.Traditional citation analysis has a variety of practical uses,such as paper search,academic evaluation,scientific citation network drawing,etc.Citation sentiment analysis is able to identify the citation content emotionally and deepen the function and use of citation analysis.In the current academic evaluation system,academic papers are an important form of academic achievement and an important evaluation standard for researchers.Academic literature writing is of special importance to each researcher.Citation sentiment analysis analyzes and discusses the citation behavior in the paper from the perspective of citation sentiment,which can help scientific researchers to understand the citation behavior while writing their own papers.On the basis of theories of scientometrics,bibliometrics and content analysis,this article constructs a solution that uses sentiment analysis,systematically combs and develops a set of implementation methods that use sentiment analysis.Then it completes citation content extraction and citation sentiment recognition with a combination of text mining,natural language processing,society network analysis,citation analysis,and other methods.At the same time,the application value of citation sentiment analysis is indicated through case analysis.The research conclusions of this paper are as follows:(1)Citation sentiment analysis is an extension of the research on citation information analysis,which promotes the further development of bibliometrics.(2)The realization of citation sentiment analysis requires the theoretical guidance of disciplines such as scientometrics and bibliometrics.It also requires the comprehensive use of natural language processing techniques,social network analysis methods,and data visualization methods.(3)Citation sentiment is the emotional tendency of the author of the citing literature,which is excavated from the cited content.Citation sentiment has a bright application prospect in the real environment. |
PDF Full Text Request