Watson Supply Chain Ideas

Use this form to submit an idea for a new product feature. The product team will review your input and provide status updates as decisions are made regarding the request.

Before you submit a new idea, please view requests that have already been submitted. If your idea has already been submitted, you can add comments or vote on the existing idea, thereby indicating your agreement with the idea. We may use this information to help prioritize development of new features.


Submit ideas for Watson Marketing and Watson Commerce products

Update B2Bi / ITXA Document Extraction to support SPEContextParams.PDToProcessData=No in conjunction with the For Each Document Service

This functionality is beneficial for B2Bi Business Process performance when document extraction results in a large number of split documents.  BP execution can be slow if many split documents are written to Process Data which can be avoided by using this parameter in conjunction with the for each document service.


Here is the IBM documentation from native SBI Document Extraction service:



Specifies whether split documents will be stored in process data or in the business process context. Valid values are:

  • Yes – Include each split document in process data (default).
  • No – Do not include each split document on process data; instead, store the information for each document in an array named SplitDocs. Retrieve documents from SplitDocs one at a time using the For Each Document service.

The PDToProcessData parameter can be used to improve performance by reducing the overhead associated with persisting large process data information.

The major performance improvement is realized at the current step (Document Extraction) and each subsequent step by persisting a much smaller process data which does not contain multiple split documents. By using the For Each Document service in conjunction with PDToProcessData = No, only the current document is on process data. this avoids repetitive writing of non-current documents. Once documents are in the SplitDocs array, they can be retrieved only by using the For Each Document service.

  • Avatar32.5fb70cce7410889e661286fd7f1897de Guest
  • May 3 2018
How will this idea be used?

This functionality would help to avoid performance impacts of writing many split documents to SBI's Process Data after document extraction.

What is your industry? Healthcare
What is the idea priority? High
DeveloperWorks ID
Link to original RFE
  • Attach files