Former Editor-in-Chief, L. Wilson Pearson, has provided the following insightful comments about keyword selection:

ScholarOne Manuscripts provides space for authors to enter keywords for their manuscripts directly. It also provides a picklist function so that authors can select words from the official list of IEEE keywords.

Please employ the picklist in choosing your keywords. Dreaming up keywords at one’s own discretion completely defeats the purpose of having keywords in the first place.

I have learned that authors largely misunderstand the purpose of keywords. They tend to use them to point to the uniqueness of the contribution of the paper in questions. We are all prone to a bit of self-aggrandizement, so this is understandable. The intent of keywords is to conclude a year with as small a set of keywords as possible to span all of the papers published in that year. Having more than one term to refer to a single concept defeats this purpose. The choice of keywords enables (or defeats) readers in searching out papers on a particular topic. Consider the following example.

L. W. Pearson and R. A. Whitaker, “A Transverse Aperture-Integral-Equation Solution for Edge Diffraction by Multiple Layers of Homogeneous Material,” Radio Science, Vol. 26, No. 1, January, 1991.

The late Rick Whitaker and I did this work in the late 1980s. We were proud of the concepts introduced in the paper and our pride might have dictated a choice of keywords along the line.

• Transverse-aperture integral equation
• Method of moments
• Dielectric slabs
• Electromagnetic diffraction
• Asymptotic anticipation

The first of these hypothetical keywords reflects author vanity and nothing more. The last term is not a part of any official keyword list, though it has been used by a number of workers. The middle three are legitimate keywords.

The difficulty with “transverse-aperture integral equation” is quite evident if you reflect on it: no one would ever think of looking under this topic in an annual index to try to find this paper. It is not at all universal. The term from the list, “integral equation methods” will do just fine in place of the more pretentious term.

Of course, keyword lists cannot be static. New concepts come into our endeavors all the time. As they become commonly accepted, then the master list of keywords must be expanded. In perspective, the IEEE list contains almost 8000 entries. (Actually, the IEEE adopts a list that is maintained by INSPEC, an arm of the IEE, London.) The editor of any transactions has the prerogative to submit extensions to the list. I maintain an expansion list that grows by a few words as each issue is indexed. I have added about 150 words in 2-1/2 years as editor. Many of these reflect omissions on the part of the INSPEC list. The following are examples.

• Dielectric slabs
• Evanescent waves
• Impedance boundary conditions
• Propagators
• Quasi optics

Some examples of terms that have recently grown common enough to add are as follows:

• Dielectric resonator antennas
• Fast solvers
• Numerical dispersion
• Urban propagation

One can clearly recognize that these terms have grown in importance to the point of common usage. The term “fast solvers” is an example worthy of note. Adding this to the index was my choice. I am sure that there is a community that would like to see “fast multipole method” indexed as a term of its own. “Fast solvers” is a broader term into which the fast multipole method fits nicely. The fact that both terms begin with the word “fast” ensures that anyone seeking the fast multipole method in an index will come upon the “fast solver” heading. My choice was simply that I elected to use the broadest possible terms.

What should an author do if he finds adequate indexing terms to be lacking for a new submission? Simply indicate the fact in the “Author Comments” field when the paper is submitted into Manuscript Central, suggesting some possibilities for a new index term.