From Bayesian optimization to extensions: How to talk about JMPAny language, including the language we use to talk about JMP, is constantly evolving. To ensure we are all consistent and correct, whether it be in promotional materials, emails, or presentations, here are a few reminders and updates.
Bayesian optimization
With the release of JMP 19, Bayesian optimization has become a hot topic for everyone at JMP. Senior Communications Manager Landra Robertson (@LandraRobertson) and I have seen enough permutations of the phrase and its abbreviations that we thought a little guidance was necessary. So here goes:
- When referring to the general concept of Bayesian optimization, only Bayesian should be capitalized. Example: I think every scientist in the world should know about Bayesian optimization.
- When referring SPECIFICALLY to the platform in JMP, then Optimization is capped: The Bayesian Optimization platform in JMP 19 is amazing.
- It should never be abbreviated when referring to the JMP platform.
- While you should avoid abbreviating Bayesian optimization in most external materials, if you need to abbreviate it, only the concept should be shortened, never the platform name. The only acceptable abbreviation is Bayes opt. Notice the lower-case o and the space between the words. Example: The words in Bayesian optimization are so long that it’s faster to type Bayes opt.
- NEVER, ever, ever abbreviate it as BO. For most native English speakers, BO refers to body odor. JMP shouldn’t be associated with anything that could have a negative connotation for some people.
Extensions
When the JMP Marketplace was created, everyone referred to its offerings as add-ins, connectors, apps, and/or extensions, using the terms interchangeably. But they aren’t all equal. If you want to use one word to describe everything, extension is the best choice. It’s the umbrella term under which all the other ones fit. The other words are still valid and applicable for specific categories of extensions, but if you need to refer to everything in the JMP Marketplace, extension is what you should use.
Design of experiments
Design of experiments is hardly a new concept, but there still seems to be some confusion when referring to it. So let me set the record straight:
- Similar to Bayesian optimization, design of experiments is lowercased when referring to the general concept. Example: Every scientist in the world should use design of experiments.
- The exception to that capitalization is if you are referring to the platform in JMP. Example: No other software has anything close to the magnificent Design of Experiments platform in JMP.
- If you want to abbreviate design of experiments, it should always be DOE, all caps, all the time. NEVER DoE.
I’ll close with a reminder that the JMP Business Communications Style Guide has all this information, as well as a lengthy list of other style conventions used by JMP. Bookmark it and refer to it before you write about JMP. Or create a PowerPoint presentation. Or post on social media.
As always, if you are confused by the Style Guide or have any questions about it, ask me.
With the release of JMP 19, we need to ensure consistency in promotional materials, emails, or presentations.
Company UpdateJMP NewsRe: From Bayesian optimization to extensions: How to talk about JMPWhen running Bayes Opt, the script that pops up in the data table says BO!

Re: From Bayesian optimization to extensions: How to talk about JMPHi @monique_lander, Please ping me on Teams and let's create a case to get that fixed in an upcoming release.
Re: From Bayesian optimization to extensions: How to talk about JMP@monique_lander -san and @PatrickGiuliano -san,
As far as I see, it has already changed to "Bayes Opt" in JMP19.1 and JMP20.0 under development.
Just for your information: This wording problem does not happen in Japanese (and maybe Simplified Chinese). We can write "Bayesian Optimization" just by only eight Japanese-Chinese characters (or 6 characters). (We can also write "Design of Experiments" by only four Japanese-Chinese characters.)
Re: From Bayesian optimization to extensions: How to talk about JMPall good thanks for the update @yusuke_ono -san.
Re: From Bayesian optimization to extensions: How to talk about JMP@yusuke_ono -san, japanese must be so simple
Thanks @AmyStuart for this clear explanation. Others might have realized as well - "BO" is occasionally used in publications, amongst our customers (e.g. Evonik, BASF, BI published "BoFire", and at least the people I speak to abbreviate with "BO").
So it's good for me to have a clear guidance on how we should use the terminology around Bayesian optimization!