Improving search keywords

[ltalirz]

This issue tracks limitations of current search terms to track citations of codes and suggestions on how to improve them.

• "xTB Grimme" is a bit too broad as it searches both for the implementation and for the method. spot checks of citing articles show that in most cases Grimme's implementation is used as well, but suggestions for a better query string are welcome
• early versions of gromacs did not list Lindahl as an author - replace by "GROMACS" Lindahl or Berendsen
• change tb-lmto-asa search term to "TB-LMTO-ASA"
• "what if" vriend contains wrong matches
• exciting results contain wrong matches. consider replacing by https://scholar.google.com/scholar?cites=1156082755018988671&as_sdt=2005&sciodt=0,5&hl=en
• "LAMMPS -Plimpton" has about half as many hits as "LAMMPS Plimpton". Spot-checks show that many of those cite lammps correctly but are missed probably because the manuscript body is not indexed (and lammps is mentioned in the title/abstract).
• Similarly, "nwchem" without any author does not seem to include any bad matches - but has significantly more (~+50%)
• "CASTEP Payne" misses many valid references. "CASTEP" (Payne OR "Materials Studio" OR "Material Studio") is better
  (note: "Material Studio" is a very common typo)
• "BOSS program Jorgensen" yields many incorrect matches
• There is Dalton and LSDalton - update the search string accordingly
• "Q-Chem" Shao includes (a minority of) incorrect matches
• "Sherrill Psi4" gives significantly less matches than citations of the 3 corresponding papers 1.0, 1.1, 1.4.
  Including first authors of these papers in the search string significantly improves agreement (" "Sherrill" OR "Turney" OR "Parrish" OR Smith "Psi4" ")
• "CASINO" "Needs" Rios still has many incorrect matches. Better use "Needs" "Quantum Monte Carlo" CASINO
• GULP Gale has a number of incorrect matches. '"GULP" Gale Program OR Software OR Rohl' substantially reduces them
• the search results for Espresso ("molecular dynamics" OR "soft matter" espresso holm) do contain a couple of "Quantum ESPRESSO" instances
• On 2022-03-20, counting references for MOLCAS/OpenMOLCAS by journal reference yield a grand total of 464 citations for 2021 (see below), compared to the 211+166=377 listed for the project. To investigate whether the query string misses citations from some references

https://www.sciencedirect.com/science/article/pii/S0927025603001095

46

https://onlinelibrary.wiley.com/doi/abs/10.1002/jcc.21318

50

https://onlinelibrary.wiley.com/doi/abs/10.1002/jcc.24221

173

https://onlinelibrary.wiley.com/doi/abs/10.1002/qua.20166

17

https://pubs.acs.org/doi/abs/10.1021/acs.jctc.9b00532

174

https://scholar.google.com/scholar?cluster=9353997742036508063&hl=en&oi=scholarr

1

https://onlinelibrary.wiley.com/doi/pdf/10.1002/wcms.1117

3

In general, put search terms in quotation marks.
Often it does not make a difference, but sometimes it does (prevents matches with different words of similar spelling)