user complains about one of the projects which goes from source language Japanese that for example one of the segments contains contains string "A/P" which is included in the Japanese terminology:
This termbase is applied to the project, so A/P should come up with a suggested term, but it doesn't. Is this because it is only three characters? Other, longer, segments are getting suggested terms.
For Asian languages the minimum length for a "non-Asian" word (ie: word with Latin letters such as "Hello", "Computer" etc.) is two characters. In the text "A/P" you have two words of one character each: "A" and "P".