Shape the future of IBM watsonx Orchestrate
Start by searching and reviewing ideas others have posted, and add a comment (private if needed), vote, or subscribe to updates on them if they matter to you.
If you can't find what you are looking for, create a new idea:
stick to one feature enhancement per idea
add as much detail as possible, including use-case, examples & screenshots (put anything confidential in Hidden details field or a private comment)
Explain business impact and timeline of project being affected
[For IBMers] Add customer/project name, details & timeline in Hidden details field or a private comment (only visible to you and the IBM product team).
This all helps to scope and prioritize your idea among many other good ones. Thank you for your feedback!
Specific links you will want to bookmark for future use
Learn more about IBM watsonx Orchestrate - Use this site to find out additional information and details about the product.
Welcome to the IBM Ideas Portal (https://www.ibm.com/ideas) - Use this site to find out additional information and details about the IBM Ideas process and statuses.
IBM Unified Ideas Portal (https://ideas.ibm.com) - Use this site to view all of your ideas, create new ideas for any IBM product, or search for ideas across all of IBM.
ideasibm@us.ibm.com - Use this email to suggest enhancements to the Ideas process or request help from IBM for submitting your Ideas.
@Guest it would be helpful to provide more information before we can decide on a course of action for this idea. For document extraction, there are many aspects that affect the overall quality of extraction.
The language of the documents
The languages supported the chosen LLM support
The language the extractor fields and schema are defined in
We have seen from limited testing that you can get good results when the language of a document is different than that of the language the extractor fields the schema is defined in.
I suggest doing some testing with sample documents in the target language(Chinese) first on IBM cloud.
The list of supported LLMs for the extractor changes over time. I suggest looking through the LLMs in the extractor dropdown to see if any of them specifically mention support for Chinese. If any of them state it explicitly, that's a good one to start with. If none of the LLMs state support for Chinese, it's worth still trying them. In many cases the LLM may have been trained on some Chinese content, but not enough for the model provider to explicitly claim support for a given language. However, depending on the documents being processed the results can still be acceptable.
Also try creating a schema in English at first, especially if the use case includes processing documents in more than one language. We have seen in testing cases where an schema defined in english does a good job extracting content from documents in a non english language. If this doesnt yield good results, you can try defining the extractor in the language of the target documents. To support this we'll soon be releasing support in the extractor UI where it allows a larger character set for the schema field input which should enable using chinese characters in the field names and descriptions.
If through testing you are able to identify a schema definition and an LLM that produces good results, please post an update here and we can then look at enabling this capability in the on-prem in a subsequent wxO release.
Note other related ideas on Chinese language support, please comment with more details, including customer/project name(s) and timeline(s) on any of those you see relevant, thanks!
additional LLMs strong with the Chinese language : https://watson-orchestrate.ideas.ibm.com/ideas/LSABER-I-999
Knowledge support : https://watson-orchestrate.ideas.ibm.com/ideas/LSABER-I-1325
Default chat welcome & prompts:
https://watson-orchestrate.ideas.ibm.com/ideas/LSABER-I-1277 for welcome message
https://watson-orchestrate.ideas.ibm.com/ideas/LSABER-I-1276 for start prompts
Gen AI: https://ibm-data-and-ai.ideas.ibm.com/ideas/LSABER-I-947 (let us know if this one is not good enough and brings us back to https://watson-orchestrate.ideas.ibm.com/ideas/LSABER-I-999)
UI translation: https://watson-orchestrate.ideas.ibm.com/ideas/LSABER-I-996
We appreciate your input, understanding the impact on ongoing or upcoming customers/projects will help prioritize those improvements, thank you!