Research On Identifying Scientometrics Hot Topics With Multiple Data

Identifying hot topics is the key to reflecting the development of a specific research field,and also the important step to make science and technology policy making evidence-based.With the abundance of scientometrics data in the era of digital publishing,the data selection that could be used to identify hot topics becomes more and more abundant.This dissertation uses metadata(publication data)and three other types of impact data(citation data,usage data and Altmetrics data)to conduct the research.At first,the definition,relationship and difference of these four categories of data are investigated.Besides,this dissertation takes 12914 papers published by five international journals in the field of scientometrics as research objects to quantitatively reveal the temporal distribution and paper distribution of four kinds of data on the basis of publication data,cited times,usage count provided by Web of Science and Altmetric Attention Score provided by altmetric.com.Based on the definition and characteristics of four types of data,this dissertation analyzes the applicability of them in identifying hot topics.Lastly,this dissertation takes scientometrics as an example to identify hot topics by four categories of data respectively and draws the co-occurrence network.This dissertation traces the change trend of hot topics identified by four kinds of data in different time periods to investigate data optimal selection from different perspectives of time and attention.And the potential of usage data and Altmetrics data in identifying hot topics are explored and proved in this dissertation.