Azure Text Analytics - Object of type 'System.String' cannot be converted to type 'System.Data.DataTable'.

[AlexDuff]

Afternoon,

I'm creating a test process that reads a scanned pdf using:

• Computer Vision: Read Raw Image

This is cast to JSON collection and results in a final collection containing 80 rows of text.

This works up to this stage and the text is saved in a collection called "Final Collection"

The issue I have is using:

I'm presented with the following:



Do I need to read the pdf into a collection prior to running the "Recognise Sentiment"? The supporting documents for these skills are limited in their explanation.

Any help or guidance on this matter would be appreciated (or any example process that can be provided in order to mimic it but alter it for my collection).

Thanks,

Update:

I've switched the value to [] instead of specifying a column within the collection.

However, I've now experienced: "Recognise Sentiment" due to an error when running custom code: Column 'Language' does not belong to table Collection_of_Documents."


The input for this action doesn't request a 'Language' to be provided. It's all wonderfully confusing and I suspect late Friday afternoon isn't a good time to be doing this.

EDIT

I've introduced another column to the collection which forces "en" as the language. This now stops the above error from happening.

However, I'm now presented with:


I'm not really sure how to resolve an operational timeout

[AlexDuff]

Updated question