smarter transformation and integration of enterprise data with Power Query and Data Integration Platform12-07-2019 0 Comment(s)
Power Query is being dramatically enhanced with market-leading Smart Data Preparationexperiences, including:
- Data extraction from semi-structured sources like PDF files.
- HTML pages.
- Fuzzy-matching algorithms to rationalize and normalize data based on similarity patterns.
- Data profiling capabilities.
These abilities help to easily identify errors and outliers as part of the data preparation workload in the Power Query Editor.
Microsoft is leveraging its leadership in the artificial intelligence field based on years of investment in several projects within Microsoft Research, and surfacing these Smart Data Preparation capabilities in Power Query, making them easily accessible to millions of business users across different Microsoft products and services (Excel, Power BI, Common Data Service for Apps, and Microsoft Flow).
Intelligent transforms and AI support in Power Query
Business analysts can easily incorporate AI-driven insights with a single-click access to AI-based transforms. Initial capabilities in this area will include sentiment analysis and keyword extraction from natural language text. Additional capabilities such as OCR and image analysis could be added over time.
Analysts and BI Pros will gain access to new out-of-the-box functions and APIs for AI transformations. This facilitates column transformations for sentiment analysis and keyword extraction with a click of a button, or with a single line of script, such as “ApplySentimentIndex([textResponse])”.
The Power Query APIs will also be supported for use during installation and configuration of Power BI applications making use of dataflows. In this instance, batch-processing of the transforms for the column will be directed seamlessly to a Cognitive Services container running as part of a customer’s Power BI Premium capacity.
No Comments to Show