Genesys Cloud - Main

Hello. 
We are using the Out of the box Topics as well as others created for our business needs. I have noticed that I need to input many different renditions of the same phrase to ensure capturing the different way phrases could be stated, yet I am also seeing phrases captured as "accurate" for a Topic that are not scripted in the way they are being captured. 

Example #1 - 
Out of Box Topic - Agent Knowledge Gap
Captured Phrase (in BOLD) - (because you) know you don't (have to get it in WiFi mode)
Topic Phrase "Matched" - I don't know
My observation - If the phrase "I don't know" will match "know you don't" this is not intuitive or helpful and will likely provide other mis-matches as well. 

Example #2 -
Out of Box Topic - Dissatisfaction
Captured phrase - (that) makes sense
Topic Phrase "Matched" - makes no sense
My observation - The captured phrase is not Dissatisfaction, it is agreement. 
​
This makes no sense to me, how phrases that mean the opposite of what is scripted are being counted as accurate. 
Thank you for any input or tips anyone can provide for more accurate phrase capturing in the calls.
#Reporting/Analytics

------------------------------
Janelle Lopez
The Chamberlain Group, Inc.
------------------------------

Hi Janelle,

Thanks for sharing these examples with us. We are improving topic spotting with customer's feedback. We are releasing a feature very soon that will give you the ability to perform an exact match when using double quotes while defining a topic. For example you can add phrases such as "makes no sense" or "no sense" to Dissatisfaction, and it will not make the mistake of capturing 'that makes sense'.

I am confident this will solve this issue you're facing.

------------------------------
Anik Dey
Genesys - Employees
------------------------------

Original Message:
Sent: 11-08-2022 15:24
From: Janelle Lopez
Subject: Topics - not working efficiently

Hello.
We are using the Out of the box Topics as well as others created for our business needs. I have noticed that I need to input many different renditions of the same phrase to ensure capturing the different way phrases could be stated, yet I am also seeing phrases captured as "accurate" for a Topic that are not scripted in the way they are being captured.

Example #1 -
Out of Box Topic - Agent Knowledge Gap
Captured Phrase (in BOLD) - (because you) know you don't (have to get it in WiFi mode)
Topic Phrase "Matched" - I don't know
My observation - If the phrase "I don't know" will match "know you don't" this is not intuitive or helpful and will likely provide other mis-matches as well.

Example #2 -
Out of Box Topic - Dissatisfaction
Captured phrase - (that) makes sense
Topic Phrase "Matched" - makes no sense
My observation - The captured phrase is not Dissatisfaction, it is agreement.
​
This makes no sense to me, how phrases that mean the opposite of what is scripted are being counted as accurate.
Thank you for any input or tips anyone can provide for more accurate phrase capturing in the calls.
#Reporting/Analytics

------------------------------
Janelle Lopez
The Chamberlain Group, Inc.
------------------------------

Hi Anik, 

You're welcome, and I thank you for your update. I look forward to the feature you mentioned and can't wait to see it implemented and test it out!

------------------------------
Janelle Lopez
The Chamberlain Group, Inc.
------------------------------

Original Message:
Sent: 11-09-2022 10:03
From: Anik Dey
Subject: Topics - not working efficiently

Hi Janelle,

Thanks for sharing these examples with us. We are improving topic spotting with customer's feedback. We are releasing a feature very soon that will give you the ability to perform an exact match when using double quotes while defining a topic. For example you can add phrases such as "makes no sense" or "no sense" to Dissatisfaction, and it will not make the mistake of capturing 'that makes sense'.

I am confident this will solve this issue you're facing.

------------------------------
Anik Dey
Genesys - Employees

Original Message:
Sent: 11-08-2022 15:24
From: Janelle Lopez
Subject: Topics - not working efficiently

Hello.
We are using the Out of the box Topics as well as others created for our business needs. I have noticed that I need to input many different renditions of the same phrase to ensure capturing the different way phrases could be stated, yet I am also seeing phrases captured as "accurate" for a Topic that are not scripted in the way they are being captured.

Example #1 -
Out of Box Topic - Agent Knowledge Gap
Captured Phrase (in BOLD) - (because you) know you don't (have to get it in WiFi mode)
Topic Phrase "Matched" - I don't know
My observation - If the phrase "I don't know" will match "know you don't" this is not intuitive or helpful and will likely provide other mis-matches as well.

Example #2 -
Out of Box Topic - Dissatisfaction
Captured phrase - (that) makes sense
Topic Phrase "Matched" - makes no sense
My observation - The captured phrase is not Dissatisfaction, it is agreement.
​
This makes no sense to me, how phrases that mean the opposite of what is scripted are being counted as accurate.
Thank you for any input or tips anyone can provide for more accurate phrase capturing in the calls.
#Reporting/Analytics

------------------------------
Janelle Lopez
The Chamberlain Group, Inc.
------------------------------