Implement Knowledge Mining and Document Intelligence Solutions¶

Overview¶

This domain covers implementing Azure Cognitive Search solutions with enrichment pipelines, and implementing Azure AI Document Intelligence for document processing. It represents 10-15% of the exam and focuses on extracting structured information from unstructured content.

Key Topics¶

Azure Cognitive Search Implementation¶

• Data source creation and configuration
• Index definition (fields, attributes, analyzers)
• Indexer creation and execution
• Index querying (syntax, sorting, filtering, wildcards)
• Knowledge Store projections (file, object, table)
• Search relevance tuning
• Faceted navigation implementation

Enrichment Pipeline¶

• Cognitive Services account attachment to skillset
• Built-in skill selection for documents
• Custom skill implementation and integration
• Skillset design and configuration
• AI enrichment for content understanding
• Skill caching for cost optimization

Azure AI Document Intelligence¶

• Prebuilt Document Intelligence models usage
• Custom Document Intelligence model implementation
• Model training, testing, and publishing
• Composed model creation
• Form field extraction
• Document structure analysis (tables, text)

Services Reference¶

Core Services¶

• Azure Cognitive Search
• Azure AI Document Intelligence (Form Recognizer)
• Azure Cognitive Services (for built-in skills)
• Azure Storage (for data and knowledge store)

Supporting Services¶

• Azure Blob Storage
• Azure Table Storage
• Azure Data Lake Storage
• Azure Functions (for custom skills)
• Azure App Service (for custom skill hosting)

Best Practices¶

Search Index Design¶

• Choose appropriate field types and attributes
• Use analyzers for language-specific processing
• Implement suggester for autocomplete
• Design for query patterns you need to support
• Use scoring profiles to tune relevance
• Consider index size and query performance

Data Source Configuration¶

• Schedule indexer runs appropriately
• Handle incremental updates efficiently
• Configure deletion detection for data source
• Monitor indexer execution and errors
• Use field mappings when needed
• Implement change tracking when available

Skillset Development¶

• Use built-in skills when possible
• Cache enrichments to reduce costs
• Design custom skills for specialized needs
• Implement error handling in custom skills
• Test skills before production deployment
• Monitor skill execution performance

Document Intelligence¶

• Choose appropriate prebuilt model when available
• Label sufficient training data for custom models
• Use consistent document formats for better accuracy
• Test model with various document variations
• Implement confidence threshold handling
• Version custom models appropriately

Common Scenarios¶

Search Applications¶

• Enterprise document search
• E-commerce product search
• Knowledge base search
• Media asset search
• Resume database search
• Legal document discovery

Knowledge Mining¶

• Document insight extraction
• Entity recognition across corpus
• Sentiment analysis of documents
• Language detection in multi-language content
• Key phrase extraction for tagging
• OCR for image-based documents

Document Processing¶

• Invoice processing and data extraction
• Receipt digitization
• Business card information extraction
• Form processing automation
• Contract analysis
• Medical record processing

Integration Patterns¶

• Upload document triggering indexing
• Search results displayed in web app
• Custom skill calling external API
• Knowledge Store for downstream analytics
• Document Intelligence in workflow automation
• Search integration with chatbot

Common Scenarios (continued)¶

Advanced Search Features¶

• Faceted navigation for filtering
• Autocomplete and suggestions
• Geo-spatial search for location-based content
• Semantic search for better relevance
• Multi-language search support
• Search result highlighting

Study Tips¶

• Practice creating search indexes with various field types
• Understand index attributes (searchable, filterable, sortable, facetable)
• Learn search query syntax and operators
• Hands-on experience with skillset creation
• Study built-in cognitive skills and their capabilities
• Practice creating custom skills with Azure Functions
• Understand knowledge store projections (shaper skill)
• Learn Document Intelligence Studio interface
• Practice training custom document models
• Study prebuilt model capabilities (invoice, receipt, ID, business card)
• Understand indexer scheduling and monitoring
• Learn about search analyzers and tokenizers
• Practice implementing scoring profiles