NLU virtual agent topic

[mehak1890]

Dear All,

I have some basic questions about NLU. I have enabled the NLU and i have created a Model/ intent then i have created the 6 utterances and also created same 6 test cases under test model and i have not created any entities under same intent.

Is it working fine or not? I'm not sure , but when I'm searching how to request the incident/ or problem on chatbot , the topic immediately starts, let's assume it's working fine. But when I'm searching another request in chatbot "I want to raise a request for phishing". The same topic "report a problem" is starting immediately, you know when I'm starting a new request then it should give me phishing topics or knowledge articles for phishing related if any in servcienow correct? which is not under the utterances what I understand. Can anyone pls help me how i can correct my utterances? so that if i'm searching " i want to raise a request for phishing" then some other results should come in chatbot?

Thanks,

Mehak

[Tricia Cornish]

Hello @mehak1890 

To get automation topics, have a specific function for your VA topic. If "phishing" is a situation you want to address as a VA topic, then your intents are going to have example utterances around that subject: "I might have a virus", "I've been phished", "I clicked on a link", "I received a scam" etc. In that VA topic, your utterances are specific to the subject "phishing". When the user types in a "phishing" related utterance, the topic will display.  Within that VA topic, you can create a specific type of ticket/task record when this topic is launched.  This "phishing" topic is not going to be used when a user types in "report a problem with my monitor" or "I need a new badge".

It might help to review "Email issues" or "Email Setup" (there's an example of this topic in the ITSM Virtual Agent Conversations and ITSM NLU Model in the content packs), you'll see that when a user types in very specific utterances around 'email" and email -related utterances (outlook, messages, mail, authentication)  this topic will surface. 
Let's say you create a "Phishing" topic and you have an "Email" topic. And then your user types in "Request a new badge" (which might be a record producer in your catalog). Neither the phishing or email topic should display. You don't need to create a new topic. Instead, I would rely on AI Search - Fallback to surface KB and Catalog content that might answer the user's question. That topic presents the user with a "did you find this helpful?" and if the user says no, then "Fallback" the topic can surface options to further assist the user. I modified my fallback topic to create a ticket. I shared this solution on a VA Academy. https://youtu.be/E1gc3JoUr6A
Hope that helps, 

Tricia 

[mehak1890]

Hi Tricia, 

Thanks for the response. 

Yes, I understand it. Also, I watched the video, but when I'm typing 'I want to request a phishing' or if I'm trying to request a pizza, the chatbot takes the "Report a problem" topic by default. As you can see, I have given "I want to request" in my utterances. Also, when I'm trying to 'give feedback,' the "Report a problem" topic is displayed automatically in the chatbot. There is no keyword in an utterance like give or feedback. I don't know why it's taking the same topic.

If I'm typing only 'phishing', it goes back to Fallback topics and gives multiple results. I tried to pass mapped entities as per the below

 

But it's also not working. I think we have to pass this statement "I want to request" under utterances and then some variable in entities that is what I understood. 

Thanks,

Mehak

[mehak1890]

Hi Tricia/ All,

Can anyone help me on this? 

Thx

Mehak

[Brian Bakker · Accepted solution]

@mehak1890 

When it triggers unwanted intents "report a problem" for certain utterances "give feedback", then this is generally caused by one of the following - 

1. No vocabulary defined for acronyms or business terms used in the NLU Model, hence the NLU Engine can return unexpected predictions because it requires a synonym that expands the acronym and its meaning.
2. Samples utterances that have no context and consist of one or two words. Ensure all utterances have full context for the intent and avoid repeating common words that are not unique to the intent. AI does not like repetition, as it will skew predictions.

I see this issue time and time again. Hence, we introduced LLM topic discovery in Virtual Agent, which does not require the creation or maintenance of any NLU model, and we also provided tools for Migrating NLU/keyword Virtual Agent topics to LLM topics.

Hope this helps.

Regards,

Brian