AI Bot
AI-200: Language & Knowledge Mining Flashcards
Microsoft Azure AI Cloud Developer Associate AI-200 Flashcards
Study our AI-200: Language & Knowledge Mining flashcards for the Microsoft Azure AI Cloud Developer Associate AI-200 exam with 50+ flashcards. View as flashcards, a searchable table, or as a fun matching game.
| Front | Back |
| Define entity in LUIS | A piece of data extracted from an utterance that provides details such as a date location or product |
| Define intent in LUIS | Represents the goal or purpose behind a user utterance used to route to appropriate action |
| Define NER | NAMED ENTITY RECOGNITION detects and classifies named entities like people organizations locations and dates |
| Describe waterfall dialog | A sequence of steps each handling a turn used to manage multi step conversational flows |
| How can you use Azure Search with QnA | Index QnA content enable semantic ranking and use search results to supplement or fallback to QnA answers |
| How do you create a QnA knowledge base | Import FAQs or documents or add pairs manually then publish the knowledge base to a host endpoint |
| How do you create embeddings for vector search | Use an embeddings model from Azure AI to convert text into numeric vectors then store them in the index |
| How do you handle profanity or PII in Text Analytics | Use content moderation filters PII detection and redact or mask sensitive fields during enrichment |
| How do you improve QnA matching quality | Add alternative phrasings configure synonyms tune threshold and use multi turn prompts |
| How do you improve utterance coverage in LUIS | Add diverse labeled utterances use phrase lists and include variations of user language |
| How do you integrate LUIS with a bot | Call LUIS during message processing to extract intent and entities then route to dialog logic or actions |
| How do you perform batch translation in Azure | Use the Translator batch endpoint or Azure Storage integration to translate large volumes asynchronously |
| How do you secure Language and Search services | Use role based access control managed identities private endpoints and encryption at rest and in transit |
| How do you test a LUIS app | Use the LUIS portal utterance testing console or call the runtime prediction endpoint with sample utterances |
| How do you version and manage LUIS models | Use application versions export and import snapshots and employ CI CD pipelines for deployment |
| What are analyzers in Azure Search | Tokenization and normalization components that control how text is broken into terms for indexing and searching |
| What are evaluation metrics for intent classification | Precision recall F1 score and confusion matrix for model performance analysis |
| What are key phrases | Important words or short expressions that summarize main points of a document or sentence |
| What is a composite entity in LUIS | An entity composed of other entities to represent hierarchical or related values |
| What is a custom analyzer | A user defined analyzer that combines tokenizers char filters and token filters to meet language needs |
| What is a data source for Azure Search | Connection configuration that defines where documents come from like blob storage or Cosmos DB |
| What is a phrase list feature in LUIS | Domain specific vocabulary that boosts recognition for specific terms or synonyms |
| What is a prebuilt entity in LUIS | An entity type provided by service for common concepts like datetime number or email |
| What is a QnA knowledge base | A collection of question and answer pairs used to provide automated responses to user questions |
| What is a scorer profile or scoring profile | A configuration that modifies result ranking using weighted fields functions and boosting rules |
| What is a skillset in Cognitive Search | A collection of cognitive skills used to enrich content during indexing for capabilities like OCR NER and language detection |
| What is a suggester in Azure Search | Component that enables type ahead and autocomplete by specifying fields to use for suggestions |
| What is adaptive dialog | Dialog style that uses declarative rules and triggers to adapt to user input dynamically |
| What is an Automated Glossary in Translator | Custom term mapping to preserve translations of domain specific terms and brand names |
| What is an enrichment step | Intermediate processing such as language detection or OCR applied to content before indexing |
| What is an index field type for vectors | Collection of single dimension floating numbers or vector field that supports approximate nearest neighbor search |
| What is an index in Azure Search | A schema that defines searchable fields types and attributes for storing documents |
| What is an indexer | Automated component that pulls data from a data source and populates an Azure Search index on a schedule |
| What is an indexer skill that calls the Text Analytics API | Custom skill that invokes Text Analytics for NER sentiment or key phrase extraction during enrichment |
| What is Azure Cognitive Search | Managed search service for indexing querying and enriching content at scale |
| What is concept of multi turn QnA | Follow up prompts and context maintained across turns to deliver multi step question answering |
| What is entity linking | Associating recognized entities to entries in a knowledge base to provide canonical identifiers and info |
| What is language detection | Identifies the language of a given text and returns a language code and confidence score |
| What is LUIS and its primary purpose | Language Understanding Intelligent Service for intent and entity extraction to drive conversational applications |
| What is metadata in a QnA pair | Key value tags added to QnA entries to help filter and route answers at query time |
| What is pattern matching intent in LUIS | Intent defined by regex or pattern rules to capture structured utterances such as order numbers |
| What is responsibility of cognitive services deployment tier | Determines SLA capacity latency and cost considerations for production or testing workloads |
| What is semantic search in Azure Cognitive Search | Search feature that uses semantic ranking and embeddings to return more relevant results beyond keyword matching |
| What is sentiment analysis | Determines positive neutral or negative sentiment and can return confidence scores |
| What is slot filling in dialog management | Process of collecting required pieces of information from user to complete an intent using prompts |
| What is synonym map | A list of equivalent terms applied at query or index time to expand matching synonyms |
| What is the Text Analytics API | Azure service for NER sentiment key phrases language detection and linked entities |
| What is Translator service | Cloud service for text translation across many languages with real time and batch options |
| What is utterance annotation in LUIS | Labeling user phrases with intents and entities to train and improve the model |
| What is vector search in Azure Cognitive Search | Capability that uses vector embeddings for semantic similarity matching of documents and queries |
About the Flashcards
Flashcards for the Microsoft Azure AI Cloud Developer Associate exam focus on core language understanding and conversational design concepts. Review intent and entity modeling (including composite and prebuilt entities), phrase lists and utterance annotation, intent pattern matching and slot filling, dialog styles like waterfall and adaptive, plus testing, evaluation metrics, versioning, and bot integration.
They also cover text analytics features such as named entity recognition, sentiment scoring, key phrase extraction, language detection, and entity linking; machine translation with automated glossaries and batch workflows; question-and-answer knowledge bases with multi-turn prompts and metadata; and cloud search topics including indexing, indexers, skillset-based enrichment, semantic and vector search, embeddings, security, deployment tiers, and handling PII and profanity.
Topics covered in this flashcard deck:
- Intents and entities
- Dialog management
- Text analytics and NER
- Machine translation and glossaries
- Q&A knowledge bases
- Search, indexing, vectors