I always feel uncomfortable when doing literature reviews and surveys. There is a feeling that I would never find all the important and relevant references. Even though it is easy to navigate in databases like Web of Science, Scopus, IEEE Xplore, there is no straightforward way of obtaining the full image of researches done in a field of interest.
I felt this issue especially acute this year when I was working on the paper “Benefits of Corrective Closed-Loop Power System Control via Combined Frequency Control and Congestion Management” together with professor Janusz Bialek. We had a hard time during the review process in IEEE Transactions on Power Systems journal. SPOILER: our paper was rejected at the second stage of the review. Among all the issues, the reviewers mentioned incompleteness of our literature review section. I felt unsatisfied with this situation and decided to do a proper literature survey.
While working on the paper, it became clear that the very first study on security-constrained optimal power flow (SCOPF) was [A. Monticelli, M. Pereira, and S. Granville, “Security-constrained optimal power flow with post-contingency corrective rescheduling,” IEEE Trans. Power Syst., vol. 2, no. 1, pp. 175–180, 1987]. However, the paper by Monticelli was cited more than 300 times, according to Scopus. It is barely possible to analyze hundreds of papers. Moreover, those papers were subsequently cited more than six thousand times. The only solution that I came up with was to implement a citation network analysis.
It was a long journey of composing the citation network. I had to use graph analysis software Gephi, Selenium web applications, and program my own scripts to download info from Scopus, track papers and citations.
I did not have time yet to finalize and publish the literature survey. But I hope that it may be useful for experts interested in SCOPF and related fields:
The interactive visualization of the network: http://materials.andreychurkin.ru/network1/
I’ve got positive feedback on the survey and the citation network analysis implemented. Several of my colleagues in Skoltech already asked for composing similar networks for their topics. I am ready to assist everyone who contacts me regarding the citation network analysis.
I hope that this approach may significantly improve research studies, especially at the early stages.
Andrey Churkin (Андрей Чуркин) 2019