- SAP Managed Tags:
Dear AI enthusiasts,
It's time to introduce you to the latest & greatest features introduced in our SAP AI Business Services portfolio this quarter! – including:
As you may already know, all our services are available on SAP Business Technology Platform (SAP BTP).
Automate and scale your business document processing for a wide range of document types using machine learning.
You have the ability to build and test custom document processing scenarios for free using templating with the free tier plan available in SAP BTP.
For more information, read this blog post.
We enable the provision of a preconfigured schema, SAP_OCROnly_schema, for optical character recognition (OCR) output only.
It speeds up processing time for documents, for which there is no need to extract SAP-managed fields (such as header fields or line items of supplier invoices):
For more information, read this blog post and watch the feature in action.
You can now achieve better accuracy for postal address segmentation when processing purchase orders.
We have improved the pre-trained model for purchase orders for certain address fields with the introduction of address segmentation based on NER for the whole address.
The user interface of Document Information Extraction is now available in the following languages depending on their location: English, German, French, Japanese, Portuguese, Russian, Simplified Chinese, Traditional Chinese, Spanish, Italian and Korean.
Did you know that Document Information Extraction capabilities are natively integrated in the new SAP Build Process Automation available on SAP Business Technology Platform (SAP BTP)?
Using the SAP BTP free tier option, you can even start in your production environment right away, try it out up to a specific capacity limit, and switch to the paid service plan smoothly, without losing your work. Benefit right away from:
Watch this video to see it in action!
Get accurate recommendations based on users’ browsing history and/or item description using machine learning.
When interaction data between user and items is not available, the new content-based approach will learn recommendations using item metadata (item catalogue, item attributes, textual features), helping you to:
Perform classification and regression tasks using machine learning.
This enhancement allows the usage of more features for a single model training.
The new maximum feature size is now 60 if the dataset contains 10 or less textual features.
This is also independent of the chosen model template.
Current business blueprints remain unaffected by this change.
Thank to SAP AI Launchpad available on SAP BTP, developers can manage their AI scenarios and models - like SAP AI Business Services - across multiple instances of AI runtimes (such as SAP AI Core).
We've recently released this step-by-step guide on how to manage Business Entity Recognition models, one of the SAP AI Business Services, from an easy to use UI with SAP AI Launchpad.
See more here! 👇
See you next year for groundbreaking AI innovations! 👍
It's time to introduce you to the latest & greatest features introduced in our SAP AI Business Services portfolio this quarter! – including:
As you may already know, all our services are available on SAP Business Technology Platform (SAP BTP).
Document Information Extraction 📲
Automate and scale your business document processing for a wide range of document types using machine learning.
Templating function available within the SAP Business Technology Platform free tier
You have the ability to build and test custom document processing scenarios for free using templating with the free tier plan available in SAP BTP.
For more information, read this blog post.
OCR-only schema
We enable the provision of a preconfigured schema, SAP_OCROnly_schema, for optical character recognition (OCR) output only.
It speeds up processing time for documents, for which there is no need to extract SAP-managed fields (such as header fields or line items of supplier invoices):
- Upload documents for OCR processing only
- Select SAP_OCROnly_schema to have the service return the full OCR result
For more information, read this blog post and watch the feature in action.
NER-based address segmentation for purchase orders
You can now achieve better accuracy for postal address segmentation when processing purchase orders.
We have improved the pre-trained model for purchase orders for certain address fields with the introduction of address segmentation based on NER for the whole address.
Internationalization of the Document Information Extraction user interface
The user interface of Document Information Extraction is now available in the following languages depending on their location: English, German, French, Japanese, Portuguese, Russian, Simplified Chinese, Traditional Chinese, Spanish, Italian and Korean.
Document Information Extraction embedded in SAP Build Process Automation
Did you know that Document Information Extraction capabilities are natively integrated in the new SAP Build Process Automation available on SAP Business Technology Platform (SAP BTP)?
Using the SAP BTP free tier option, you can even start in your production environment right away, try it out up to a specific capacity limit, and switch to the paid service plan smoothly, without losing your work. Benefit right away from:
- Data extraction – Extract data from large numbers of digital documents or scanned images and transfer it to your enterprise systems for processing.
- Structured and unstructured documents – Extract data from structured data sources such as Microsoft Excel files and unstructured data sources such as PDF, PNG, TIFF, JPEG, and more.
- Data enrichment – Enrich extracted data based on your own master data and connect your incoming document-based extractions to a unique ID that your system understands to process further.
Watch this video to see it in action!
Personalized Recommendation 📲
Get accurate recommendations based on users’ browsing history and/or item description using machine learning.
Content-based recommendations to enhance onboarding experience
When interaction data between user and items is not available, the new content-based approach will learn recommendations using item metadata (item catalogue, item attributes, textual features), helping you to:
- Overcome the lack of an AI training data set and support AI model development
- Enhance the AI recommendation model to deliver higher business value and new revenue streams
- Reduce privacy risks
Data Attribute Recommendation 📲
Perform classification and regression tasks using machine learning.
Enhanced feature selection
This enhancement allows the usage of more features for a single model training.
The new maximum feature size is now 60 if the dataset contains 10 or less textual features.
This is also independent of the chosen model template.
Current business blueprints remain unaffected by this change.
SAP AI Business Services + SAP AI Launchpad = Perfect Match
Thank to SAP AI Launchpad available on SAP BTP, developers can manage their AI scenarios and models - like SAP AI Business Services - across multiple instances of AI runtimes (such as SAP AI Core).
We've recently released this step-by-step guide on how to manage Business Entity Recognition models, one of the SAP AI Business Services, from an easy to use UI with SAP AI Launchpad.
See more here! 👇
See you next year for groundbreaking AI innovations! 👍
