Presents a working, fully calibrated model of citation dynamics that is ready to use
Offers a quantitative example of how concepts developed in the field of complex networks help to solve the real-life problem of modeling the citation dynamics of papers
Demystifies the origin of the power-law statistical distribution of citations
Michael Golosovsky is an experimental physicist and he has been doing research and teaching physics in the Hebrew University of Jerusalem since 1988. He published more than 100 papers in the peer-reviewed journals in the fields of solid state physics, biophysics, and complex networks. During last decade he focused his attention on citation networks and brought to this interdisciplinary field his expertise in planning and performing measurements. Basing on these measurements, he succeeded in building a physical, data-based model of citation dynamics.
This book deals with the science of science by applying network science methods to citation networks and uniquely presents a physics-inspired model of citation dynamics. This stochastic model of citation dynamics is based on a well-known copying or recursive search mechanism. The measurements covered in this text yield parameters of the model and reveal that citation dynamics of scientific papers is not linear, as was previously assumed. This nonlinearity has far-reaching consequences including non-stationary citation distributions, diverging citation trajectories of similar papers, and runaways or "immortal papers" with an infinite citation lifespan. The author shows us that nonlinear stochastic models of citation dynamics can be the basis for a quantitative probabilistic prediction of citation dynamics of individual papers and of the overall journal impact factor. This book appeals to students and researchers from differing subject areas working in network science and bibliometrics.
I: Introduction.
A: The place of citation analysis in science: bibliometrics, power-law distributions, complex networks.
B: The purpose of the book: a quantitative (calibrated and verified) model of citation dynamics.
II: Complex network of scientific papers.
A: Network structure.
B: Synchronous and diachronous citation distributions.
C: Reference-citation duality.
II: Age distribution of references in the reference lists of scientific papers.
A: The author´s strategies to compose the reference list of a paper.
B: Recursive search algorithm: direct and indirect references.
C: Modeling age distribution of references.
D: Comparison to measurements and model calibration.
III: Citation dynamics of scientific papers- a mean-field model.
A: A mean-field model of citation dynamics of individual papers.
B: Comparison to measurements and model calibration.
1. Direct citations.
2. Indirect citations.
3. Statistics of the second-generation citations and citing papers.
4. Probability of indirect citation.
C: Analysis of the mean-field model.
1. Citation trajectories of individual papers.
2. Citation lifetime and runaway papers.
IV: Discrete stochastic model of citation dynamics of individual papers.
A: Model formulation.
B: Model verification.
1. Citation distributions.
2. Citation trajectories: general shape, fluctuations, autocorrelation.
3. Statistics of uncited papers.
V: Prediction of citation dynamics of individual papers.
A: Citation trajectory.
B: Paper´s fitness and impact factor.
VI: ”Power-law” citation distributions.
A: Time-resolved citation distributions and their approximation by the power-law dependences.
B: Numerical simulation using our model: citation distributions are non-stationary.
C: Power-law citation distribution is a transient phenomenon arising from nonlinear citation dynamics.
VII: Comparison of our model of the growing citation network to existing models.
A: Survey of models of the growing complex networks.
B: Equivalence between preferential attachment and fitness models.
C: Our model reproduces preferential attachment.
D: The ”genuine” preferential attachment exists and is related to nonlinear citation dynamics.
VIII: Generalization of our results for other complex networks.
A: Discrete stochastic recursive search model for growing complex networks.
B: Growth of complex networks and its relation to network structure: local assortativity and clustering coefficient.
IX: Summary
Chapter1: Introduction.- Chapter2: Complex network of scientific papers.- Chapter3: Stochastic modeling of references and citations.- Chapter4: Citation dynamics of individual papers -model calibration.- Chapter5: Model validation.- Chapter6: Comparison of citation dynamics for different disciplines.- Chapter7: Prediction of citation dynamics of individual papers.- Chapter8: Power-law citation distributions are not scale-free.- Chapter9: Comparison to existing models
Michael Golosovsky is an experimental physicist and he has been doing research and teaching physics in the Hebrew University of Jerusalem since 1988. He published more than 100 papers in the peer-reviewed journals in the fields of solid state physics, biophysics, and complex networks. During last decade he focused his attention on citation networks and brought to this interdisciplinary field his expertise in planning and performing measurements. Basing on these measurements, he succeeded in building a physical, data-based model of citation dynamics.
Über den Autor
Michael Golosovsky is an experimental physicist and he has been doing research and teaching physics in the Hebrew University of Jerusalem since 1988. He published more than 100 papers in the peer-reviewed journals in the fields of solid state physics, biophysics, and complex networks. During last decade he focused his attention on citation networks and brought to this interdisciplinary field his expertise in planning and performing measurements. Basing on these measurements, he succeeded in building a physical, data-based model of citation dynamics.
Inhaltsverzeichnis
Chapter1: Introduction.- Chapter2: Complex network of scientific papers.- Chapter3: Stochastic modeling of references and citations.- Chapter4: Citation dynamics of individual papers -model calibration.- Chapter5: Model validation.- Chapter6: Comparison of citation dynamics for different disciplines.- Chapter7: Prediction of citation dynamics of individual papers.- Chapter8: Power-law citation distributions are not scale-free.- Chapter9: Comparison to existing models
Klappentext
This book deals with the science of science by applying network science methods to citation networks and uniquely presents a physics-inspired model of citation dynamics. This stochastic model of citation dynamics is based on a well-known copying or recursive search mechanism. The measurements covered in this text yield parameters of the model and reveal that citation dynamics of scientific papers is not linear, as was previously assumed. This nonlinearity has far-reaching consequences including non-stationary citation distributions, diverging citation trajectories of similar papers, and runaways or "immortal papers" with an infinite citation lifespan. The author shows us that nonlinear stochastic models of citation dynamics can be the basis for a quantitative probabilistic prediction of citation dynamics of individual papers and of the overall journal impact factor. This book appeals to students and researchers from differing subject areas working in network science and bibliometrics.
Presents a working, fully calibrated model of citation dynamics that is ready to use
Offers a quantitative example of how concepts developed in the field of complex networks help to solve the real-life problem of modeling the citation dynamics of papers
Demystifies the origin of the power-law statistical distribution of citations
