Cloud Development Kit


 Hello. I have been having difficulty getting the subset knowledge feature to work. I uploaded a unicode file to the Myscript tool to generate the .sk file. I then added it on my account and it appears like so:


My JSON request is the following:
  "type": "start",
  "textParameter": {
  "language": "en_US",
  "textInputMode": "CURSIVE",
  "resultDetail": "WORD",
  "subsetKnowledges": ["tnum"],
  "textProperties": {
  "wordCandidateListSize": 1,
  "inputUnits": [
  "textInputType": "MULTI_LINE_TEXT",
  "components": [
  "type": "stroke",
  "t": [

  "userId": redacted,
  "containerId": "form1",
  "containerType": "FORM"

The recognition works fine without the "subsetKnowledges": ["tnum"], line added, but when I add that line in, it gives me the error: {"type":"error","error":{"result":{"language":"en","resourceName":"tnum","error":"SKResourceNotFoundException","status":"error"}}}

I've tried deleting and re-adding the reource, I've double checked that it was being generated properly. Any ideas why it might not be working?

1 Comment

Dear Remi,

thank you for contacting us.

Currently, rather than calling your custom sk resources in the "subsetKnowledges" parameters, you should instead call these using the "userResources" parameter, as follows:

"resultDetail": "WORD",
  "userResources": ["tnum"],
  "textProperties": {

Best regards,



Login or Signup to post a comment