# Heuristics for key term mismatch

**URL:** <https://forum.xbench.net/t/heuristics-for-key-term-mismatch/485>\
**Category:** General Discussion\
**Created:** [April 3, 2019, 10:20am UTC](https://forum.xbench.net/t/heuristics-for-key-term-mismatch/485 "2019-04-03T10:20:26Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![agnus](https://avatars.discourse-cdn.com/v4/letter/a/4bbf92/32.png) [@agnus](https://forum.xbench.net/u/agnus)\
**Post date:** [April 3, 2019, 10:20am UTC](https://forum.xbench.net/t/heuristics-for-key-term-mismatch/485/1 "2019-04-03T10:20:26Z")

</div>

Hi,

would that be possible to add some heuristics to key term mismatch check?

The way it works now makes it quite useless for e.g. Slavic languages:

Key Term Mismatch (iPod touch / iPod touch)  
Learn what to do if your iPhone, iPad or iPod touch is lost or stolen  
Dowiedz się, co zrobić w przypadku zgubienia lub kradzieży telefonu iPhone, iPada lub **iPoda touch**

Key Term Mismatch (Apple Support / Wsparcie Apple)  
If you need additional help, contact Apple Support.  
Jeśli nadal potrzebujesz pomocy, skontaktuj się ze **Wsparciem Apple**.

It should be self-explanatory where the problem lies…

---

<div class="post-metadata">

**Author:** ![RSchiaffino](https://yyz1.discourse-cdn.com/flex035/user_avatar/forum.xbench.net/rschiaffino/32/15_2.png) [@RSchiaffino](https://forum.xbench.net/u/RSchiaffino)\
**Post date:** [April 4, 2019, 5:05pm UTC](https://forum.xbench.net/t/heuristics-for-key-term-mismatch/485/2 "2019-04-04T17:05:02Z")

</div>

Two possible solutions:

1. Use the key term glossary, but enter in it only the root part for the target language, so for example  
Support \<tab\> Wsparci
2. Instead of using the key term glossary, add the terms to a personal or project checklist, where you can use regex to account for the different possible word endings, e.g.  
Apple Support Wsparci(e|em) Apple  
(This of course might need to be changed depending on how your language actually works… I often do this for Italian, for example so as to be able to use the same check for both singular and plural forms of a term, e.g.: “programm(a|i) di assistenza Apple”)

---

<div class="post-metadata">

**Author:** ![agnus](https://avatars.discourse-cdn.com/v4/letter/a/4bbf92/32.png) [@agnus](https://forum.xbench.net/u/agnus)\
**Post date:** [April 4, 2019, 10:57pm UTC](https://forum.xbench.net/t/heuristics-for-key-term-mismatch/485/3 "2019-04-04T22:57:31Z")

</div>

I am aware of these workarounds. They will not work at all in my workflows, in which I run Xbench over Studio projects with attached termbase. In such case, the keyterm file (from attached Multiterm) will be each time loaded from new file, which unables me to do any persisent changes in the keyterm file or to move these checks to the checklist.

Hence my heuristics request.