AI-900 Question 22
Select 3A company wants to build a customer support chatbot that can understand user queries, respond with relevant information, and escalate complex issues to a human agent when necessary. Which of the following workloads are involved in implementing this solution?
- A
Text classification to categorize user intents
- B
Entity recognition to extract important details from user queries
- C
Speech-to-text conversion to transcribe audio messages
- D
Image recognition to analyze customer-uploaded documents
- E
Sentiment analysis to detect the emotional tone of user messages
Show answer and explanation
Correct answers: A, B, E
Explanation
The scenario involves a natural language processing (NLP) workload where a chatbot needs to understand user queries, extract important information, and assess the emotional tone to escalate issues as needed. The correct workloads include text classification for identifying intents, entity recognition for extracting details, and sentiment analysis for understanding user emotions. Speech-to-text and image recognition are not applicable in this context.
- A. Correct.
Text classification is essential for identifying the intent behind user queries, enabling the chatbot to respond appropriately.
- B. Correct.
Entity recognition is used to extract specific details, such as names, dates, or order numbers, from user input, which helps the chatbot provide personalized and accurate responses.
- C. Incorrect.
Speech-to-text conversion is not relevant in this scenario since the question does not mention handling audio input.
- D. Incorrect.
Image recognition is unrelated to the described chatbot functionality, as there is no mention of analyzing customer-uploaded documents in the scenario.
- E. Correct.
Sentiment analysis is useful for understanding the emotional tone of user messages, which can help the chatbot determine when to escalate issues to a human agent.