Training
Identify and quickly train unrecognized user inputs in your NLX workspace
Last updated
Identify and quickly train unrecognized user inputs in your NLX workspace
Last updated
Training lists all user utterances, whether spoken or written, that your deployed bots were unable to match to an intent. Training allows you to match unknown phrases to the appropriate intent, refine existing training phrases to better adapt to user language, or to identify possible actions or intent flows your uses require.
The Unknown tab lists all utterances or inputs from users that were unrecognized by NLP and were unable to match to an intent with confidence.
Unknown messages are organized by the following:
Text: The user input or utterance
Sent: Date the input or utterance was made
Confidence score: The confidence level of the NLP in matching the user's utterance
Intent: The intent that the NLP routed the user to
Trained intent: If trained, displays the intent the user input was added to
Unknown messages remain in the list, even if training has occurred. Trained messages appear with information under the Trained intent column.
Clusters group user utterances semantically, providing you with a visual presentation of key topics commonly queried by your users. They can provide insight into the needs of users as well as opportunities for automation:
Toggle between Table or Map view to change the visual presentation of captured clusters. Clusters provide the following information:
Label: The semantic description provided to the cluster group (can be renamed by typing in the Label field in Table view)
Phrases: The number of user inputs (with total percentage) that comprise the cluster
Unknowns: The number of times the user input triggered the bot's Unknown response
Intents: The number of intents where user inputs matching the semantic cluster appeared
Subclusters (only visible in Table view): The number of individual sub groups that comprise the cluster. Expand each cluster row to view
Hover over a row and select the Expand icon (in Table view) or select the card (in Map view) to view advanced details, including all intents and timestamp of phrases.
Notice associations between clusters? Easily merge clusters that you'd like grouped together. Simply select each cluster and choose the Merge option.
You may use any filter to narrow the results by Bot, Intent, Region, Language, Start & End times, Channel, exact text (Search), and/or Confidence score. Results automatically update as filters are applied.
You may easily add an unknown user input as a training phrase to an intent to better train the NLP model.
Select Training from workspace menu > Click the checkbox to the right of the message you wish to train
Choose Train intent link. You can select as many items as you'd like, but they must all be added to the same intent, as only one intent's training data may be improved at a time
Select an intent from the dropdown
Select/deselect language(s) associated with the chosen intent
Enter any phrase adjustments to alter the inputs (messages will otherwise be added to an intent's list of Training phrase as they appear)
Click Confirm
Trained messages appear with information under the Trained intent column and as part of an intent's Training phrases.
Be sure to create a new build and re-deploy your bot to experience any changes.
