Research On The Impact Factors Of Highly Cited Papers Based On Web Of Science Large-scale Literature Data

With the exponential growth of academic data and the development of big data technologies,there are many established methods of academic data acquisition and pre-processing available in existing research.Scholars from different interdisciplinary fields have conducted a large number of studies on research publications,and have produced many theories and methods that have underpinned scientific research.Highly cited papers are widely regarded as potentially "excellent" scientific research and are being attached more and more importance to in scientific evaluation.The study of laws of highly cited papers is conducive to a better understanding of the internal structure of science and is a guide to the conduct of scientific research activities.Recent researches on highly cited papers have mainly explored the correlation between highly cited papers and such certain factors as interdisciplinarity,reviews,international collaboration,length of reference list,etc.,exploring whether those factors can increase the number of citations a paper receives and thus make it more likely to become a highly cited paper.In addition,there is a part of works based on highly cited papers to evaluate scientific work at the national,institutional,journal and other levels.However,due to the difficulty of obtaining big academic data and it is difficult to deal with and analyze in the traditional library and information field,most of the studies on highly cited papers have certain limitations,such as small amount of data and low persuasive power,and usually do not take into account domestic and international citation,there is no research on the impact of domestic citation and international citation on highly cited papers.Therefore,this thesis obtains more than 30 million papers and more than 500 million citation relationship data on the web of science platform,by data preprocessing,represents the country,year and other information in the form of dictionary,and the citation relationship is represented by CSR adjacent sparse matrix,so it can analyze large-scale literature data more efficiently,and selects the top 1% of highly cited papers according to the year of publication,and analyzes the impact factors of highly cited papers by country in three aspects: the impact of domestic citation on highly cited papers,the impact of international citation on highly cited papers,and the citation extent among different countries:(1)The impact of domestic citation on highly cited papers in 12 countries,including China.By 2012,China was already second only to the United States in the number of highly cited papers.However,excluding domestic citations,China's relative world ranking in terms of the number of international highly cited papers would be significantly lower.Through four different similarity coefficients,this thesis quantifies the consistency of highly cited papers with international highly cited papers across countries.The results show that the number of Chinese highly cited papers is significantly increased by domestic citations,but not by other countries.At the same time,by further analyzing the differences between Chinese highly cited papers and international highly cited papers,this thesis finds that the impact of domestic citations on Chinese highly cited papers is mainly in the field of chemistry.(2)The impact of citations between the three countries or consortium of the United States,Europe,and China on highly cited papers.When all citations between Europe and the United States are excluded,the number of highly cited papers around the world fluctuates more significantly,while when all citations between Europe or the United States and China are excluded,there is little fluctuation in the number of highly cited papers in other countries except for the country to which the citation is removed,leading to a phenomenon that implies that there may be citation aggregation between European and American countries.In addition,China has gradually gained the upper hand in the Sino-American-European citation competition.If China,the US,and Europe do not cite each other,the US and Europe will lose more highly cited papers than China.But there may be a hidden phenomenon that is not good for China: China is increasingly inclined to cite literature from Europe and the United States,while Europe and the United States are increasingly less inclined to cite Chinese literature.(3)To better explain the issues raised in(2),based on the zero model,this thesis proposes a method to quantify the citation extent among different countries.First,this thesis builds a ciation heat map based on the z-score matrix,and analyzed the citation aggregation of different countries.It was found that there were two significant societies at the citation level,European,American countries and Asian countries,with higher citation aggregation within societies.Then,the pattern and trends of the degree of citation in different countries between 2007 and 2017 were analyzed.The analysis shows that China is increasingly inclined to cite papers from European and American countries,while most countries,including European and American countries,are increasingly less inclined to cite papers from China.Based on the journal partition data from the JCR of Chinese Academy of Sciences,this thesis investigates the citation extent of high-level papers in China by different countries and finds that the overall conclusion is relatively stable,the citation extent of Chinese papers in different countries is still on the decline.Only by correctly understanding the objective laws in scientific research and the internal structure of science,can we promote the development of science faster.The study of the objective laws of highly cited papers is of great significance.On the one hand,the study of the impact factors of highly cited papers is conducive to the formulation of government research lines and research evaluation policies,and also serves as a good reference for government research incentive policies.On the other hand,the study of the impact factors of highly cited papers can play a guiding role in the training of researchers,and can provide a reference for researchers to choose the research direction.