Document processing capabilities
The document processing capabilities of Microsoft Syntex include prebuilt document processing models for contracts, invoices, and receipts. These models are pretrained to process common business use cases and for structured documents. Prebuilt models use OCR and deep learning models to extract data from structured documents. Prebuilt models are a good method to start automating document processing and content extraction.
Custom document processing models support structured, freeform, and unstructured documents. The difference between these models is the training method. The structured model uses the layout method where the model creator trains the model by marking the content fields from the document. On the freeform model, the model is trained by selecting the content for extraction anywhere on the document. On unstructured documents, the model is trained by teaching the model. Freeform and structured models are created and trained using the Microsoft...
