aholanda / k-index

K-index applied to Web of Science.

K-INDEX

• Aim: Use K-index as a predictor of who will be awarded with a scientific prize.
• Testbed: gather most of the Highly Cited Researchers in Physics from Clarivate and try to predict the Nobel Laureates. The use of K-index is not limited to Nobel Laureates prediction. We compare the results with the standard scientometric index used by Web of Science, the h-index (Hirsch index).
• Manuscript: arXiv:1910.02369, 2019.

Research Team

• Osame Kinouchi;
• Adriano J. Holanda;
• George C. Cardoso.

Reproducibility

The results can be totally reproduced by performing the following instructions:

Automatic

To reproduce the experiment, clone the repository, install make, cweb and a C compiler and run

$ make

Manual

The following steps are less automated but they have less dependencies:

• Download the data using the link;
• Unzip the downloaded data;
• Download the program binary Linux-x86_64 or Windows-x86_64 according to the platform;
• Run the program in the same directory where the data were extracted.

Documentation

• Program documentation.

Contributor

• Luis Fernando Castro - fixes some grammatical and spelling errors.

References

• J. E. Hirsch. "An index to quantify an individual's scientific research output," PNAS 102 (15) 16569-16572, 2005. [Wikpedia entry]
• O. Kinouchi, L. D. H. Soares, G. C. Cardoso. "A simple impact index for scientific innovation and recognition," arXiv:1609.05273v2, 2017.

k-index's Issues

Add verbosity to program output.

Some information about the data is stored inside the files. The enhancement required is the addition of verbosity flag to write that information to output.

Remove Nobel winners.

Find a mechanism to remove researchers from physics that already won the Nobel prize.

Add 3 authors of Web of Science prediction to win the Nobel

Add the following 3 researchers in the list. They are the bet of Web of Science to win the Nobel in Physics:

1. Artur K. Ekert, Professor of Quantum Physics, Mathematical Institute, University of Oxford, UK, and Lee Kong Chian Centennial Professor, National University of Singapore, Singapore.For contributions to quantum computation and quantum cryptography. Recognized for fundamental research that unites theoretical and experimental physics with computer and information science. He is the inventor of entanglement-based quantum cryptography.
2. Tony F. Heinz, Professor of Applied Physics and Photon Science, Stanford University, and Associate Laboratory Director for Energy Sciences, SLAC National Accelerator Laboratory, Stanford, CA, United States.For pioneering research on optical and electronic properties of two-dimensional nanomaterials. We recognize Heinz for contributions to understanding classes of nanoscale materials including carbon nanotubes, graphene, and two-dimensional semiconductors such as molybdenum disulfide.
3. John P. Perdew, Laura H. Carnell Professor of Physics and Chemistry, Department of Physics, Temple University, Philadelphia, PA, United States.For advances in density functional theory of electronic structure, revealing ‘nature’s glue’. For contributions to more thorough understanding of the nature and behavior of materials. Density functional theory provides electronic structure calculations in condensed matter physics and quantum chemistry and predicts, for example, the energy of atomic bonds.

Twelve larger Ks

Generate a procedure to write as output the twelve researchers in descending order according to K index.

Fix some wrong phrases in program documentation.

The program documentation needs a review to remove some errors. The main source of errors are typos and some reasoning phrases not so correct.

Fix incorrect output of some h and K indices.

Some researchers like Zhenan Bao are showed with low number of h and K, e.g. h = 13 e K = 25. However Bao has h = 115 according to Google Scholar. Fix this error of counting.

Add more Nobel laureated researchers in the laureated_out.dat

A short initial list was set in the laureated.dat, that list can be expanded to avoid future generation of k-nobel.md with already laureated researchers.

Remove code to calculate h-index.

The h-index is being retrieved from Web of Science
and not calculated from CSV files. So the code to calculate it can be removed from k-nobel.w.

Produce a h versus K plot

The program must produce a plot h versus K to insert in the paper. In the plot the twelve larger Ks must be marked somehow.

Convert CWEB to R