Using smart mapping

Smart Mapping uses anonymized, consent-based community patterns to deliver accurate recommendations by maintaining strict privacy protection.

Smart Mapping recommends field mappings by using machine learning algorithms that learn from your previous mappings and similar patterns across integrations. It analyzes field names, data types, and hierarchies to deliver accurate, context-aware suggestions.

You can adjust mapping confidence levels (high, medium, low) and optionally use anonymized community data to improve prediction accuracy. All shared data is aggregated and cleansed of business-specific and personally identifiable information before processing.

Key capabilities of Smart Mapping
  • Community data integration - Aggregated mapping patterns enrich prediction models, business-specific, and user-specific data is not considered.
  • Better recommendations - Suggestions adapt to evolving usage patterns.
  • Privacy-first design - Data usage complies with governance policies.
  • Consent-based participation - Mapping insights are shared only with explicit consent.
Benefits of Smart Mapping
  • Faster integration development - Reduces manual mapping effort by learning from your mappings and community patterns.
  • Better, adaptive suggestions - Improves over time as more data is processed, delivering recommendations that evolve with usage trends.
  • Confidence control - Supports switching between high, medium, and low confidence levels for mapping decisions.
  • Accuracy made smarter - Combines your experience with community insights for better outcomes.
  • Business data logs
  • Static values that you set in the pipeline map
  • Accounts
  • Reference data
  • Pipeline data
  • Certificates
  • User details
  • Static values in any application operations

Operation flow of Smart Mapping

Smart Mapping operates through a privacy-preserving, pattern-based learning model:

  1. Learning from your mappings - The algorithm applies machine learning techniques to examine user-defined mappings and deliver similar recommendations for future integrations.
  2. Confidence levels - The recommendations are categorized into high, medium, and low confidence levels, enabling selection based on accuracy requirements.
  3. Community data - If enabled, Smart Mapping analyzes anonymized mapping patterns from the webMethods Integration community to enhance prediction accuracy.
    • All shared data is aggregated and stripped of business-specific and personally identifiable information before processing.
    • Patterns include structural relationships such as field names, data types, and hierarchies.
  4. Context-aware suggestions - The common mapping behaviors across integrations are analyzed to deliver accurate, context-aware recommendations that speed up development.
Important: Smart Mapping provides recommendations only. Review and validate these suggestions before you accept them, and make corrections if any mappings do not fit your specific use case.

Consent and opt-out

Opting out of Smart Mapping is allowed for users with the appropriate role permissions. When Smart Mapping is disabled, mapping recommendations are provided based on your mappings, but you do not benefit from community intelligence.

Data privacy and scope

  • Data that is collected during the opt-in period is stored in the data center that is provisioned for the user.
  • All community data that is used for predictions are anonymized and aggregated across the community and are not tenant-specific
  • Business-specific and personally identifiable information is removed before processing.
  • Only structural relationships (field names, data types, hierarchies) are analyzed to maintain privacy.
  • The following information is not indexed in flow services and workflows:
    • Business Data log
    • Static values that you set in the pipeline map
    • Accounts
    • Reference Data
    • Pipeline Data
    • Certificates
    • User Details
    • Static values in any application operations