Microsoft

AdminLUIS UserVoice (Software Engineer, Microsoft Cognitive Services)

My feedback

  1. 2 votes
    Sign in
    (thinking…)
    Sign in with: Facebook Google
    Signed in as (Sign out)

    We’ll send you updates on this idea

    An error occurred while saving the comment

    Hello,
    Currently there are no short-term plans to expand on language support, we will revisit the roadmap in January 2021.

  2. 1 vote
    Sign in
    (thinking…)
    Sign in with: Facebook Google
    Signed in as (Sign out)

    We’ll send you updates on this idea

    0 comments  ·  Language Understanding (LUIS) » Sentiment Analysis  ·  Flag idea as inappropriate…  ·  Admin →
  3. 1 vote
    Sign in
    (thinking…)
    Sign in with: Facebook Google
    Signed in as (Sign out)

    We’ll send you updates on this idea

    0 comments  ·  QnA Maker » API  ·  Flag idea as inappropriate…  ·  Admin →
  4. 9 votes
    Sign in
    (thinking…)
    Sign in with: Facebook Google
    Signed in as (Sign out)

    We’ll send you updates on this idea

    4 comments  ·  Language Understanding (LUIS) » Utterances  ·  Flag idea as inappropriate…  ·  Admin →
    Completed  ·  Luke Bayler responded

    Hello,

    Punctuation no longer has any effect in training or predicting.

    Thanks,
    Luke

    An error occurred while saving the comment

    Hmm, I would think an AI system that accounts for all input (or as much input as possible) is more powerful than an AI system that doesn't. Remember that LUIS operates using confidence scores, so it already is able to recognize questions without question marks. It's just that when a question mark is present, LUIS is more confident about the intent than if a question mark is not present, and that makes sense.

    Consider ambiguous input. If a user says "And I'm supposed to click the button?" that's an entirely different intent from "And I'm supposed to click the button." One is asking for clarification and the other is a statement of understanding.

    However, it is possible to achieve the functionality you're looking for on the client application's side. You could strip question marks off the end of utterances before they're sent to LUIS and that would achieve the effect of making sure an utterance returns the same intent regardless of whether there's a question mark. (Of course, LUIS is nondeterministic by nature so it's best not to expect the exact same confidence scores even when the exact same utterance is used twice.)

  5. 2 votes
    Sign in
    (thinking…)
    Sign in with: Facebook Google
    Signed in as (Sign out)

    We’ll send you updates on this idea

    1 comment  ·  QnA Maker » API  ·  Flag idea as inappropriate…  ·  Admin →
  6. 3 votes
    Sign in
    (thinking…)
    Sign in with: Facebook Google
    Signed in as (Sign out)

    We’ll send you updates on this idea

    Completed  ·  1 comment  ·  Language Understanding (LUIS) » Language Support  ·  Flag idea as inappropriate…  ·  Admin →
    An error occurred while saving the comment

    Thank you for your suggestion! However, LUIS already supports the Korean language. If you are asking for support for prebuilt entities in Korean; please visit another suggestion and cast your vote there.

    https://cognitive.uservoice.com/forums/551524-luis/suggestions/32293282-add-prebuilt-entites-for-korean

  7. 35 votes
    Sign in
    (thinking…)
    Sign in with: Facebook Google
    Signed in as (Sign out)

    We’ll send you updates on this idea

    An error occurred while saving the comment

    Thank you for the suggestion and your comment! Currently, datetimeV2 supports DD/MM format when it is unambiguous as to which value is the day, and which value is the month. Meaning, if the text contains "26/01", the resolution for this value would be "XXXX-01-26", January 26th. However, the prebuilt reverts to a en-US datetime format when the format is ambiguous, meaning if the text contains "03/12", this would resolve to "XXXX-03-12", or March 12th, as opposed to "XXXX-12-03", December 3rd.

  8. 1 vote
    Sign in
    (thinking…)
    Sign in with: Facebook Google
    Signed in as (Sign out)

    We’ll send you updates on this idea

    Completed  ·  1 comment  ·  Language Understanding (LUIS) » Entities  ·  Flag idea as inappropriate…  ·  Admin →
    An error occurred while saving the comment

    Thanks for the feedback Peter! We actually do support timespan entity extraction through the use of the prebuilt datetimeV2 entity. In the response from LUIS, the type extracted would be
    "builtin.datetimeV2.duration". Considering the tokens, "3 years", when 3 years is extracted, an example response would contain a "timex" value: "P3Y", and a "value" in seconds, in this case, "94608000".

  9. 5 votes
    Sign in
    (thinking…)
    Sign in with: Facebook Google
    Signed in as (Sign out)

    We’ll send you updates on this idea

    2 comments  ·  Language Understanding (LUIS) » Bot Framework  ·  Flag idea as inappropriate…  ·  Admin →
    Declined  ·  Luke Bayler responded

    Hello,

    LUIS does not currently support dialogs.

    Thanks,
    Luke

    An error occurred while saving the comment

    We have deprecated the action feature, or action binding from LUIS itself. In April our awesome documentation team put together three blog posts that show how to implement LUIS Action Binding on the client side. The blog posts are as follows:

    - https://blog.botframework.com/2017/04/03/luis-action-binding-bot/

    - https://blog.botframework.com/2017/04/06/luis-action-binding-web/

    - https://blog.botframework.com/2017/04/18/luis-action-binding-console/

  10. 5 votes
    Sign in
    (thinking…)
    Sign in with: Facebook Google
    Signed in as (Sign out)

    We’ll send you updates on this idea

    1 comment  ·  Language Understanding (LUIS) » luis.ai WebApp  ·  Flag idea as inappropriate…  ·  Admin →
    Resolved  ·  Luke Bayler responded

    Hello,

    We cannot reproduce this. If you are still having this issue, please open a ticket and send us the JSON file.

    Thanks,
    Luke

    An error occurred while saving the comment

    Is your import still failing? The actual import limit is 15 MB, so the size shouldn't be the cause of the issues. In the past, some model discrepancies (e.g. mismatched casing on entities, List Entities in excess of the 20,000 limit, duplicate entity extractor names), were determined to be the root cause of the problems you mentioned above.

Feedback and Knowledge Base