AI-102 Question 320
Select 2You are designing a multi-turn conversation bot using Azure Bot Service and Language Understanding (LUIS). The bot needs to collect user preferences for booking a flight, such as departure city, destination city, and travel dates. You want to ensure that the bot can handle scenarios where the user provides partial information and follow up with clarifying questions. What should you do to implement this functionality?
- A
Configure LUIS intents to trigger prompts for missing information.
- B
Define dialog flows with WaterfallDialog to guide the conversation step by step.
- C
Use QnA Maker to handle multi-turn conversations and store user preferences.
- D
Utilize adaptive cards to dynamically update missing information during the conversation.
- E
Implement slot filling in LUIS to automatically infer missing information.
Show answer and explanation
Correct answers: B, E
Explanation
To implement a multi-turn conversation for collecting user preferences in a flight booking scenario, you need to define dialog flows using WaterfallDialog to guide the conversation step by step. Slot filling in LUIS complements this by automating the detection of missing entities and triggering prompts for clarification. Together, these approaches enable a robust multi-turn conversational experience.
- A. Incorrect.
LUIS intents are used to classify the user’s input but do not inherently manage multi-turn conversation or missing information. Additional logic is needed in the bot framework to prompt for clarification.
- B. Correct.
WaterfallDialog is a suitable approach for defining multi-turn dialog flows. It allows you to design a step-by-step conversation where you can ask for and store specific pieces of user input such as departure city or travel dates.
- C. Incorrect.
QnA Maker is primarily designed for FAQ-style interactions and is not suitable for handling dynamic, multi-turn conversations that require collecting and managing multiple pieces of information.
- D. Incorrect.
Adaptive cards are useful for presenting information or capturing user input in a visually rich format, but they do not inherently manage the logic for multi-turn conversations or missing information.
- E. Correct.
Slot filling in LUIS allows you to define the required entities (e.g., departure city, destination city, travel dates) and automatically detect missing ones. When an entity is missing, you can configure the bot to prompt the user for the missing data.