Microsoft Azure AI Fundamentals (AI-900)Describe features of Natural Language Processing (NLP) workloads on AzureMedium
A customer support center transcribes all incoming calls for quality assurance. They need to quickly identify the primary topics of discussion within each call transcript to categorize them, even if the agents and customers use varied phrasing. The goal is to get a high-level understanding of what each call was about without reading the entire transcript. Which Azure AI NLP capability is most suitable for this?
- ANamed Entity Recognition (NER)
- BText Summarization
- CTopic Modeling
- DSentiment Analysis
Show answer & explanationAnswer & explanation
Correct answer: C. Topic Modeling
Topic Modeling is ideal for discovering abstract 'topics' (themes) that occur in a collection of documents, such as call transcripts. It can group similar calls together based on underlying themes without needing predefined categories, providing a high-level understanding.
Why the other options are wrong
- A. NER extracts specific entities like names, not the overall topic of a conversation.
- B. Text Summarization condenses a single document, but Topic Modeling helps identify common themes across many documents for categorization.
- D. Sentiment Analysis determines emotional tone, not the thematic content.
Topic Modeling
An unsupervised machine learning technique that analyzes a collection of documents to discover the abstract 'topics' that occur in them, based on word co-occurrence patterns.
- Identifies themes without pre-defined categories.
- Useful for organizing and understanding large text datasets.
- Commonly used for document analysis and content discovery.
Memory trick: Topics Model Talk for Timely Tagging