Is it possible to change QnA maker underlying AI/ML algorithms?

1
votes

Is there any way to customize qna maker in order to understand semantic phrases in a sentence?

Questions from my KB are:

  1. What is the company address?
    • Alternative question is: Company headoffice?

when user ask the "company head office?" question (Note:space between head and office) qna maker responds with some other answer which contains head office term in it.

ex:

Q. Company head office?
A: the head office has all the facilities......

Here qna maker matches the head office term from the answer and written the response.

So is there any possible way to change underlying Ml algorithms.

1
qnamaker
Do you have a specific answer you want to return for each of those questions, or is QnA Maker correctly picking the best possible answer for both cases which happens to be the same one? If the two appropriate answers do exist in your knowledge base, are the phrases "Where is reserve bank?" and "What is river bank ?" provided as alternate phrasings for those QnA pairs? – Kyle Delaney
Thanks for the quick response @KyleDelaney !! Yes, I have specific answer for both the question. Suppose if i ask what is river bank QnA maker gives response for reserve bank question and for reserve bank it provides correct answer. So here the qna maker doesn't bother about context and semantics, it looks for the best match and returns the response! so i thought if I customize algorithm which understand the semantics that might give better results. Correct me If i'm wrong. – Yeshwanth_Mandla
Are you saying "what is river bank" is an exact question in your knowledge base and QnA Maker doesn't return the answer from that QnA pair? Perhaps it will help if you edit this Stack Overflow question to provide some of the sample data from your knowledge base and sample queries along with their responses so we can see what's happening – Kyle Delaney
@KyleDelaney Sorry for the late reply! I have done some changes to my question and provided sample data from my KB. Hope it will help. – Yeshwanth_Mandla
Is Rajesh's answer acceptable? – Kyle Delaney

1 Answers

0
votes

Can we change the underlying ML algorithm for QnA Maker, sorry the answer is "No". QnAMaker is a prebuilt model offered as a service, which does not support any customization.

QnAMaker uses Natural Language Processing (NLP) and Azure Search layered ranking over the KnowledgeBase which is stored in the blob storage. In the example provided "Head Office" will be matched with the facilities related response which is expected behavior due to the higher score on ranking, but you can try adding synonyms using alterations to improve the ranking for your scenario.