AI-102 Question 321
Select 3You are designing a multi-turn conversation chatbot for a customer service application using Azure Bot Services and Azure Language Understanding (LUIS). The chatbot needs to assist users with order tracking, where it should ask for an order number if the user hasn't provided one and confirm the order details after the user provides the number. Which of the following steps are necessary to implement this functionality?
- A
Define intents and entities in LUIS to extract the user's intent and order number.
- B
Configure the bot to use Azure Cognitive Search for validating the order number.
- C
Implement dialog flows in the Bot Framework SDK to handle multi-turn conversations.
- D
Set up a QnA Maker knowledge base for storing order-related FAQs.
- E
Use state management in the Bot Framework SDK to track the user's responses across turns.
Show answer and explanation
Correct answers: A, C, E
Explanation
To create a multi-turn conversation for order tracking, you need to define intents and entities in LUIS to understand the user's intent and extract specific information like the order number. Dialog flows in the Bot Framework SDK manage the flow of the conversation, including asking follow-up questions. State management ensures the bot can keep track of user input across multiple turns. Other services, like Azure Cognitive Search or QnA Maker, are not relevant for this specific scenario.
- A. Correct.
Correct: Defining intents and entities in LUIS is essential to identify the user's purpose (e.g., order tracking) and to extract specific details like the order number.
- B. Incorrect.
Incorrect: Azure Cognitive Search is not directly relevant for validating order numbers in this scenario. This task is typically handled by a backend system or API.
- C. Correct.
Correct: Implementing dialog flows in the Bot Framework SDK is necessary for guiding the user through multi-turn interactions such as asking for and confirming the order number.
- D. Incorrect.
Incorrect: While QnA Maker is useful for FAQ-type scenarios, it is not required for managing multi-turn conversations or extracting user-provided data like order numbers.
- E. Correct.
Correct: State management is crucial in multi-turn conversations to maintain context, such as remembering the user's input (e.g., the order number) across conversation turns.