Missing training presage training data for certain languages
Bug #1321324 reported by
Michael Sheldon
This bug affects 1 person
| Affects | Status | Importance | Assigned to | Milestone | |
|---|---|---|---|---|---|
| ubuntu-keyboard |
Confirmed
|
Medium
|
Michael Sheldon | ||
| ubuntu-keyboard (Ubuntu) |
New
|
Undecided
|
Unassigned | ||
Bug Description
A number of languages currently just have empty text files used for training presage's predictive text learning model. The effected locales are: ar, cs, da, fi, he, hu, nl, ru, sv
This results in them not being able to produce good predictions until after a period of learning from user input. Appropriate source texts need to be found to replace the empty text files.
| description: | updated |
Revision history for this message
| Bill Filler (bfiller) wrote | #1 |
| Changed in ubuntu-keyboard: | |
| status: | New → Confirmed |
| importance: | Undecided → Medium |
| assignee: | nobody → Michael Sheldon (michael-sheldon) |
To post a comment you must log in.
I think we need to find ebooks or text in those languages that are copyright free and then we can use.