What is a Chatbot User Utterance?
Bu yardım dokümanının Türkçe versiyonu için tıklayın.
A user utterance is any input directed to a chatbot by a user, which can be a sentence, question, or even an emoji. These inputs are crucial for the chatbot to understand and respond appropriately.
Examples:
"How's the weather today?"
"When will my order arrive?"
"Find the nearest restaurant."
Importance of User Utterances
User utterances are essential for a chatbot to provide accurate responses and meet user needs. Properly analyzed utterances enhance user experience and improve chatbot performance. They are the fundamental components of intents defined within a chatbot.
Types of User Utterances:
Information Request: User seeks specific information.
Example: "What events are happening this weekend?"
Action Request: User requests assistance with an action.
Example: "I want to book an appointment."
Sentiment Expression: User shares emotional feedback.
Example: "I am satisfied with this service."
Problem Report: User reports an issue and seeks a solution.
Example: "I can't log into my account."
Chit-Chat: General conversation with the chatbot.
Example: "How are you?"
Processing User Utterances
Chatbots utilize Natural Language Processing (NLP) techniques to understand user utterances:
Pre-processing: Clean unnecessary words and symbols.
Intent Recognition: Identify the main purpose of the user’s input.
Entity Recognition: Detect important details (e.g., dates, locations, names).
Response Generation: Generate appropriate responses based on the recognized intent and entities.
User utterances are the cornerstone of chatbot technology. Developing an effective chatbot requires accurately processing and understanding these inputs. This guide provides an overview of user utterances, their importance, and their types, helping chatbot developers and users grasp this critical concept better.
How to Create User Utterances
To add user utterances to an intent in the Jetlink panel:
Log in to the panel.
Click on the JetBot button in the left-hand menu.
Select Edit Content on the new page.
5. Add the utterance in the displayed window, press Enter, and save.
6. Alternatively, use the Train with LLM feature:
7. Enter commands to generate utterances.
9.Click Show Suggestions.
10. Save the generated utterances to the intent.
Key Considerations for Creating User Utterances
Variation:
Create different lengths and structures for each intent.
Vary words and phrase lengths.
Change the position of entities (start, middle, end).
Mix grammar, plural forms, stems, and punctuation.
Poorly Formed Utterances:
Include examples with typos and informal language, e.g., "When will my order be delivered?" vs. "When wil my order b delivred??"
Representative Terms and References:
Use terms and references familiar to your users. For example, "Good morning" might be written as ‘’gm", ‘’gmorning’’.
Example Utterances for Each Intent:
Assign at least 50-100 sample utterances per intent.