During a chat session, the virtual assistant may not find an answer to the user's query, either because it doesn't have that content, or because it doesn't interpret the way it is placed. In such cases, the bot will respond with an evasive message. For example: "I'm sorry, I don't have that information, I promise to study to give an answer next time".
But what happens when the user does not write a valid message? In that scenario, the understanding engine identifies meaningless queries and recognizes them as disposable or "garbage".
For example, if the user randomly types "lsdjslfajsñfgfka" or "IJ4tQ92rJ", "11111111111545454", "@@@@???? " or EMOJIS, i.e. emojis that have not yet been uploaded to the content ( see our article on uploading emoji content through marketplace), the virtual assistant identifies that it is invalid content and displays a different message than the evasions for this type of interactions: "Sorry, what you wrote is not a valid word, could you express it in another way?" These evasions are called “garbage".
It is important to remember that it is not possible to add plug-ins in the garbage.
Click here to view the default customized evasions.
Important: In the statistics, these interactions will be counted as a percentage of "unresolved", but will not appear in the Training section or be considered for the transfer by consecutive evasions to human agent chat.
If you want to edit the avoidance message for invalid interaction, please request the change to our technical support via the Aivo Support Center.
Date of update: 24/08/2022