locked
How do we disable autocorrection or best guess on an ink manager in XAML c# 8.1

    Question

  • Hi

    WE are developing a metro app that is using canvas and ink recognition. As it is a spelling app we need to turn off autocorrect for the ink recognition.

    How do we disable autocorrection or best guess on an ink manager in XAML c# 8.1

    Sunday, June 29, 2014 9:09 AM

Answers

  • If you can do your own preanalysis of the cursive strokes to separate the letters then you can pass each letter to the reco engine separately. I haven't tried doing this so I don't know how successful this will be with simplistic analysis.

    --Rob

    • Marked as answer by demografix Wednesday, July 9, 2014 6:11 AM
    Tuesday, July 8, 2014 11:39 PM
    Moderator

All replies

  • At what point does the ink recognition try to autocorrect?

    Matt Small - Microsoft Escalation Engineer - Forum Moderator
    If my reply answers your question, please mark this post as answered.

    NOTE: If I ask for code, please provide something that I can drop directly into a project and run (including XAML), or an actual application project. I'm trying to help a lot of people, so I don't have time to figure out weird snippets with undefined objects and unknown namespaces.

    Monday, June 30, 2014 7:15 PM
    Moderator
  • Hi Matt

    When it returns us the recognised result. 

    A Youtube of our issue....
    https://www.youtube.com/watch?v=WfMdNceoVtM

    We are using RecognizeAsync, the get Results on an Windows.UI.Input.Inking.InkManager

    This is the sample code we have used to patch into our app.
    http://code.msdn.microsoft.com/Windows-81-Input-Ink-baf6b5e9
    Tuesday, July 1, 2014 6:13 AM
  • Correct me if I'm wrong, but it seems you're doing recognition on the entire word, not individual letters. You will probably need to analyze individual strokes to pick up letters rather than entire words. This seems a bit difficult, and it might require you to do something like put individual letters in their own boxes as some letters require multiple strokes.

    I'll also ask our inking expert to check this out.

    Matt Small - Microsoft Escalation Engineer - Forum Moderator
    If my reply answers your question, please mark this post as answered.

    NOTE: If I ask for code, please provide something that I can drop directly into a project and run (including XAML), or an actual application project. I'm trying to help a lot of people, so I don't have time to figure out weird snippets with undefined objects and unknown namespaces.


    Tuesday, July 1, 2014 1:55 PM
    Moderator
  • Thanks Matt, however individual boxes will not suit when students are writing in cursive.
    We really need to find a solution to turn auto correct off completely.
    Wednesday, July 2, 2014 10:05 AM
  • I'll see what resources we have on doing this.

    Matt Small - Microsoft Escalation Engineer - Forum Moderator
    If my reply answers your question, please mark this post as answered.

    NOTE: If I ask for code, please provide something that I can drop directly into a project and run (including XAML), or an actual application project. I'm trying to help a lot of people, so I don't have time to figure out weird snippets with undefined objects and unknown namespaces.

    Wednesday, July 2, 2014 12:00 PM
    Moderator
  • There isn't a built in way to break this down like that. The reco engine doesn't split off single letters but works with full words. If you can separate the strokes into letters yourself and then pass just the specific strokes to RecognizeAsync then you may be able to approach the behavior you are looking for.

     --Rob

    Monday, July 7, 2014 9:34 PM
    Moderator
  • Hi Rob

    So individual letter boxes and then build the word from the individual letters.
    Mmmm, so no cursive handwriting?

    Thanks

    Paul

    Tuesday, July 8, 2014 4:45 AM
  • If you can do your own preanalysis of the cursive strokes to separate the letters then you can pass each letter to the reco engine separately. I haven't tried doing this so I don't know how successful this will be with simplistic analysis.

    --Rob

    • Marked as answer by demografix Wednesday, July 9, 2014 6:11 AM
    Tuesday, July 8, 2014 11:39 PM
    Moderator
  • Mmmm... lots of cases to cover.
    Nope, I think we will go with individual letter boxes.
    Shame that the API does not have an option t simply turn off auto-correct.

    Thanks

    Paul

    Wednesday, July 9, 2014 6:10 AM