Skip to main content

Lemmatization & Multi-Conjugation Highlighting

Feature Request

The Problem: Currently, FluentAI treats different conjugations of the same verb as entirely unique entities. For example, if the word "optreedt" is used, the app does not recognize it as being related to "optreed" or "optreden."

The Suggestion: Implement Lemmatization (root word recognition) for supported languages.

  • Smart Highlighting: When a user interacts with or highlights a specific verb form (e.g., optreedt), the app should automatically highlight all other variations of that root verb (e.g., optreed, optreden) appearing in the current subtitle track.

  • Unified Word Stats: If a user "marks a word as known," all conjugated forms of that lemma should ideally be marked or linked, rather than requiring the user to "learn" every single tense or person-ending separately.

Thanks!

Posted by Ehsan ·
2 comments

Log in to comment and vote

Comments2

  • Sasha

    Team•

    Feb 8

    Wait but there should be lemmatisation, when you add a form of a verb, it is supposed to consider the base word as bookmarked, and thus all its variants

    • Ehsan

      •

      Feb 11

      Thanks Sasha for your reply! Indeed! But sometimes it doesn’t work. Look at these pictures:

      I already have “Verborgen” in the notebook which its base form is “Verbergen”. But as you can see in the subtitle, “Verbergen” was not detected!

      The same as “gevaarlijk” here:

      As you can see “gevaarlijke” is already there in the notebook and its base form was correctly mentioned “gevaarlijk”. But in the subtitle it’s not detected.