Dialog Flow Analysis Notebook in Cloud Pak For Data/Watson Studio
Have you found that you are receiving feedback from users and it is marked not helpful? Do you find that users are escalating to a live agent more frequently than they should but you are not sure why? Did my user complete the task completely or did they get frustrated and abandon the conversation ?
We have a notebook used to glean insight into issues within your dialog and how to improve the effectiveness of your virtual agent.
It uses transcript log data to identify dialog nodes and entities that commonly cause ineffective conversations. The effectiveness of a virtual assistant is more than whether or not the assistant can contain(handle the conversation without escalation to an agent) but can it complete the tasks successfully.
Demo of Python notebook: https://vimeo.com/492951672
Other Helpful Links:
Github :
Blog:
Libs:
