The term "deEss" or "de-ess" refers to editing an audio recording - usually of a voice - when the "S" (and sometimes "sh" and "th," etc.) sounds (technically called "sibilance") seem much too loud in relation to to rest of the vocal. It involves trying to reduce the volume or energy of JUST the S sound while leaving the rest of the vocal alone.
There is a specific kind of audio editing tool for this, called a "DeEsser." See my article How to Fix SSS-Sibilance in Your Audio With Sound Editing Software for the basics on sibilance and deEssers.
The way these work is that a compressor is set to only respond when certain frequencies are detected. Usually these are between 4 KHz and 8 KHz. When they ARE detected, the compressor kicks in and reduces the level of the those frequencies by a certain amount but only above a certain level. Basically it's just a frequency-specific compressor.
But there are other ways to tame those sibilant sounds. Below is a post from Audio Geek Zine with some ideas on several different ways to do it. This article is extra cool since it uses Reaper for the examples. Our course, The Newbies Guide To Audio Recording Awesomeness 2: Pro Recording With Reaper, shows you how to use Reaper to record and produce professional audio projects. I used Reaper in that course because that is my go-to recording software. The concepts taught, however, are certainly just as applicable to any other recording program.
Anyway, I digress. The article I mentioned about different ways to de-ess (deEss, whatever) a vocal is here: