AI-900 Question 30
Single answerA company has thousands of documents stored in different formats, including PDFs, Word files, and scanned images. They need a solution to extract key information such as customer names, invoice numbers, and purchase details to create a searchable database. Which of the following workloads is most suitable for this scenario?
- A
Predictive analytics to forecast customer purchase behavior
- B
Knowledge mining using Azure Cognitive Search
- C
Real-time sentiment analysis using Azure Machine Learning
- D
Building conversational AI with Azure Bot Service
Show answer and explanation
Correct answer: B
Explanation
Knowledge mining with Azure Cognitive Search is the best workload for scenarios where large volumes of unstructured or semi-structured data need to be processed, enriched, and made searchable. It leverages AI-powered capabilities like OCR, natural language processing, and entity recognition to extract and index information from diverse data sources, making it suitable for the company's requirement to process documents and create a searchable database.
- A. Incorrect.
Predictive analytics focuses on analyzing data to forecast future outcomes and trends. It is not designed for extracting and indexing information from unstructured documents.
- B. Correct.
Knowledge mining using Azure Cognitive Search is specifically designed to extract information, enrich data, and create a searchable knowledge base from various data formats, making it the ideal solution for the described scenario.
- C. Incorrect.
Real-time sentiment analysis is used to analyze emotions or opinions in text, such as social media posts or feedback. It does not address the need for extracting structured information from documents.
- D. Incorrect.
Building conversational AI focuses on creating chatbots or virtual agents for user interaction. This is unrelated to extracting and indexing data from documents.