Citation Curves Research Articles

Breakthrough potential of emerging research topics based on citation diffusion features

This article uses the characteristics of citation curves in emerging research topics (ERTs) and combines them with the ERTs’ knowledge bases to draw conclusions by comparing their development patterns. The goal of this study is to enrich the toolset for predicting breakthroughs in scientific research. A set of multidimensional and practical bibliometric indicators is used to identify ERTs, to further identify the knowledge bases of ERTs and construct citation curves for both ERTs and their knowledge bases. The development trends of the citation curves of ERTs and their knowledge bases in different time periods are compared and analysed from two dimensions: knowledge transition and continuous growth. We use the field of stem cell research to test our method. Based on the outcome of the analysis, we can assess the breakthrough potential of ERTs. The stratification, transition and recent changes of the citation curve can be used as a basis for analysing and assessing the ERTs’ breakthrough potential. The combination of different citation diffusion patterns of ERTs and their knowledge bases can improve the effectiveness of identifying ERTs that can become breakthrough innovations.

Describing Citations as a Function of Time

Abstract Purpose Providing an overview of types of citation curves. Design/methodology/approach The terms citation curves or citation graphs are made explicit. Findings A framework for the study of diachronous (and synchronous) citation curves is proposed. Research limitations No new practical applications are given. Practical implications This short note about citation curves will help readers to make the optimal choice for their applications. Originality/value A new scheme for the meaning of the term “citation curve” is designed.

Analysing the variation tendencies of the numbers of yearly citations for sleeping beauties in science by using derivative analysis

To comprehensively characterize the citation histories of sleeping beauties (SBs), this paper presents a derivative analysis of SB citation curves. Derivative analysis can differentiate among periods with different variation tendencies (ascending, declining or unchanged) in the number of yearly citations; these variation tendencies can be identified from successive (positive, negative or zero) derivatives. To overcome the interference caused by fluctuations in the citation curves, a smoothing method is first applied. A sleeping period appears in a citation curve as a horizontal region with a low citation count. During an awakening period, the derivatives of the curve are positive and may include a few zeros. Some SBs experience a rapid increase in yearly citations once awakened and have a short awakening period, whereas for others, the number of yearly citations increases steadily over a long awakening period. Different SBs can also show different forms based on the change in the number of yearly citations after they reach their peak, and these forms can also be distinguished through derivative analysis. Derivative analysis can be used alone to identify SBs and determine their awakening times or in combination with other methods of identifying SBs to improve their performance by assisting in the identification of abnormal SBs. Derivative analysis enables the “beauty coefficient’’ method (Ke et al. in Proc Natl Acad Sci USA 112(24):7426–7431, 2015) to determine awakening times that do not vary over time, thereby making Ke et al.’s method immune to interference due to citation fluctuations. It also allows one to determine multiple awakening times for a single SB.

Measuring delayed recognition for papers: Uneven weighted summation and total citations

Abstract This paper studies the so-called abnormal phenomenon of delayed recognition in bibliometrics and focuses on the first step in quantitatively measuring this phenomenon. As bibliometric analysis of a paper’s recognition and influence is an uncertain and extended process, proper calculation of delayed recognition and “sleeping beauty” publications has limitations in current scientometric studies, such as restricted application indicators, scope, and complex calculation methods. This study suggests a solution for depicting the citation delay phenomenon of individual papers that avoids dividing them into different periods, is applicable to all papers with various types of citation curves, and is easy to calculate. Notably, this approach advocates using an uneven weighted summation based on earlier and later citation years when analyzing an individual paper’s citation data. It demonstrates that the intrinsic relation between two independent indicators of citation delay and Gs index is based on the same logic of applying uneven weights to sum up yearly citations. This paper also recommends that simultaneous application of the new indicator Da and final citation numbers can efficiently identify those delayed recognition papers, and that the criterion for selecting papers can be adjusted by the value of a.

A vector for measuring obsolescence of scientific articles

Diachronous studies of obsolescence categorized articles into three general types: "flashes in the pan", "sleeping beauties" and "normal articles". These studies used either quartiles or averages to define thresholds on sleeping and awakening periods. However, such average- and quartile-based criteria, sometimes, are less effective in distinguishing "flashes in the pan" and "sleeping beauties" from normal articles due to the arbitrariness of the manner in which thresholds are determined. In this investigation, we propose a vector for measuring obsolescence of scientific articles as an alternative to these criteria. The obsolescence vector is designed as O = (Gs, A?), with Gs as a parameter affecting the shape of citation curves and A? as a parameter detecting drastic fluctuation of citation curves. We collected 50,789 articles authored by Nobel laureates during 1900---2012. Applying our criteria to this dataset, we compared the obsolescence vector with average- and quartile-based criteria. Our findings show that the proposed obsolescence vector is different from and serves as an alternative to the average- and quartile-based criteria.

Scientific research measures

The evaluation of scientific research is crucial for both the academic community and society as a whole. Numerous bibliometric indices have been proposed for the ranking of research performance, mainly on an ad hoc basis. We introduce the novel class of Scientific Research Measures (SRMs) to rank scientists' research performance and provide a rigorous theoretical foundation for these measures. In contrast to many bibliometric indices, SRMs take into account the whole citation curve of the scientist, offer appealing structural properties, allow a finer ranking of scientists, correspond to specific features of different disciplines, research areas and seniorities, and include several bibliometric indices as special cases. Thus SRMs result in more accurate rankings than ad hoc bibliometric indices. We also introduce the further general class of Dual SRMs that reflect the “value” of journals and permit the ranking of research institutions based on theoretically sound criteria, which has been a central theme in the scientific community over recent decades. An empirical application to the citation curves of 173 finance scholars shows that SRMs can be easily calibrated to actual citation curves and generate different authors' rankings than those produced by seven traditional bibliometric indices.

Sleeping beauties in genius work: When were they awakened?

“Genius work,” proposed by Avramescu, refers to scientific articles whose citations grow exponentially in an extended period, for example, over 50 years. Such articles were defined as “sleeping beauties” by van Raan, who quantitatively studied the phenomenon of delayed recognition. However, the criteria adopted by van Raan at times are not applicable and may confer recognition prematurely. To revise such deficiencies, this paper proposes two new criteria, which are applicable (but not limited) to exponential citation curves. We searched for genius work among articles of Nobel Prize laureates during the period of 1901–2012 on the Web of Science, finding 25 articles of genius work out of 21,438 papers including 10 (by van Raan's criteria) sleeping beauties and 15 nonsleeping‐beauties. By our new criteria, two findings were obtained through empirical analysis: (a) the awakening periods for genius work depend on the increase rate b in the exponential function, and (b) lower b leads to a longer sleeping period.

Citation curves of “all-elements-sleeping-beauties”: “flash in the pan” first and then “delayed recognition”

"Delayed recognition" refers to the phenomenon where papers did not achieve recognition in terms of citations until some years after their original publication. A paper with delayed recognition was termed a "sleeping beauty": a princess sleeps (goes unnoticed) for a long time and then, almost suddenly, is awakened (receives a lot of citations) by a prince (another article). There are a sleeping period and an awakening period in the definition of a "sleeping beauty". Apart from and prior to the two periods, an awaking period was found in citation curves of some publications, "sleeping beauties" was hence expanded to "all-elements-sleeping-beauties". The opposite effect of "delayed recognition" was described as "flash in the pan": documents that were noticed immediately after publication but did not seem to have a lasting impact. In this work, we briefly discussed the citation curves of two remarkable "all-elements-sleeping-beauties". We found they appeared "flash in the pan" first and then "delayed recognition". We also found happy endings of sleeping beauties and princes, and hence suggest the citation curve of an "all-elements-sleeping-beauty" include an awaking period, a sleeping period, an awakening period and a happy ending.

Group‐based trajectory modeling (GBTM) of citations in scholarly literature: Dynamic qualities of “transient” and “sticky knowledge claims”

Group‐based trajectory modeling (GBTM) is applied to the citation curves of articles in six journals and to all citable items in a single field of science (virology, 24 journals) to distinguish among the developmental trajectories in subpopulations. Can citation patterns of highly‐cited papers be distinguished in an early phase as “fast‐breaking” papers? Can “late bloomers” or “sleeping beauties” be identified? Most interesting, we find differences between “sticky knowledge claims” that continue to be cited more than 10 years after publication and “transient knowledge claims” that show a decay pattern after reaching a peak within a few years. Only papers following the trajectory of a “sticky knowledge claim” can be expected to have a sustained impact. These findings raise questions about indicators of “excellence” that use aggregated citation rates after 2 or 3 years (e.g., impact factors). Because aggregated citation curves can also be composites of the two patterns, fifth‐order polynomials (with four bending points) are needed to capture citation curves precisely. For the journals under study, the most frequently cited groups were furthermore much smaller than 10%. Although GBTM has proved a useful method for investigating differences among citation trajectories, the methodology does not allow us to define a percentage of highly cited papers inductively across different fields and journals. Using multinomial logistic regression, we conclude that predictor variables such as journal names, number of authors, etc., do not affect the stickiness of knowledge claims in terms of citations but only the levels of aggregated citations (which are field‐specific).

Scientific Research Measures

The evaluation of scientific research is crucial for both the academic community and society as a whole. Numerous bibliometric indices have been proposed for the ranking of research performance, mainly on an ad hoc basis. We now introduce the novel class of Scientific Research Measures (SRMs) to rank scientists’ research performance and provide a rigorous theoretical foundation for these measures. In contrast to many bibliometric indices, SRMs take into account the whole citation curve of the scientist, offer appealing structural properties, allow a finer ranking of scientists, correspond to specific features of different disciplines, research areas and seniorities, and include several bibliometric indices as special cases. Thus SRMs result in more accurate rankings than ad hoc bibliometric indices. We also introduce the further general class of Dual SRMs that reflect the “value” of journals and permit the ranking of research institutions based on theoretically sound criteria, which has been a central theme in the scientific community over recent decades. An empirical application to the citation curves of 173 finance scholars shows that SRMs can be easily calibrated to actual citation curves and generate different authors’ rankings than those produced by seven traditional bibliometric indices.The appendices for this paper are available at the following URL: http://ssrn.com/abstract=2535257

An extension of the h index that covers the tail and the top of the citation curve and allows ranking researchers with similar h

Citation curves for researchers with the same h index can vary greatly in the heaviness of their top (excess citations to core papers) or the heaviness of their tail (citations to non-core papers), revealing quantitative differences across researchers. Also, promotion to the next higher h depends only on citations received by a small subset of papers, so that researchers with a given h may have citation curves whose top and tail reveal a weaker impact than that of researchers with a lower h. To overcome these problems, we propose a two-sided h index, an extension that computes additional h indices progressively up the top and out the tail of the citation curve. This extension represents a citation curve descriptor one of whose elements is the scalar h. The advantages of the two-sided h index are illustrated through analysis of citation curves for 88 researchers with h indices ranging from 8 to 20. Several schemes are also discussed that use the two-sided h index to define criteria for ranking researchers within and across scalar h indices, according to whether the top of the citation curve, its tail, or both are deemed relevant under the circumstances in which research accomplishments are assessed.

Integrated impact indicators compared with impact factors: An alternative research design with policy implications

AbstractIn bibliometrics, the association of “impact” with central‐tendency statistics is mistaken. Impacts add up, and citation curves therefore should be integrated instead of averaged. For example, the journals MIS Quarterly and Journal of the American Society for Information Science and Technology differ by a factor of 2 in terms of their respective impact factors (IF), but the journal with the lower IF has the higher impact. Using percentile ranks (e.g., top‐1%, top‐10%, etc.), an Integrated Impact Indicator (I3) can be based on integration of the citation curves, but after normalization of the citation curves to the same scale. The results across document sets can be compared as percentages of the total impact of a reference set. Total number of citations, however, should not be used instead because the shape of the citation curves is then not appreciated. I3 can be applied to any document set and any citation window. The results of the integration (summation) are fully decomposable in terms of journals or institutional units such as nations, universities, and so on because percentile ranks are determined at the paper level. In this study, we first compare I3 with IFs for the journals in two Institute for Scientific Information subject categories (“Information Science & Library Science” and “Multidisciplinary Sciences”). The library and information science set is additionally decomposed in terms of nations. Policy implications of this possible paradigm shift in citation impact analysis are specified.

Characteristics of highly cited papers

Highly cited articles are very different from ‘ordinary’ cited articles. Typically, they are authored by a large number of scientists, often involving international collaboration. The majority of the papers represent regular journal articles (81%), although review articles (12%) are over-represented compared to the national average. The citation curves of highly cited papers follow a typical pattern of rise and decline. However, different types of citation curves can be identified, reflecting possible differences in the cognitive function of the articles. Highly cited papers typically obtain citations from a large number of different journals and from papers representing both close and remote fields. However, this pattern is not very different from the average distribution for all papers. We discuss how the findings can be explained by introducing a conceptual distinction between quality dynamics and visibility dynamics.

Mathematical relations between impact factors and average number of citations

Instead of the two-year impact factor as used in the Journal Citation Reports, there is much m favor of using x-year impact factors ( x>0). These impact factors are studied as a function of x and compared with the average number of citations per paper to papers that appeared in the journal x years ago. It is shown that both are equal if and only if the derivative of the impact-factor function is zero. Based on this, a simple classification of impact-factor curves versus mean citation curves is established and examples are given. These results are also applied to recent practical data that were obtained by Rousseau.