09-07-2017 04:14 AM - edited 09-07-2017 04:16 AM
I am trying to use text analytics in VA.
I have a dataset that includes 50 docs, each in a row, and I get the message that it failed because it requires 25 docs and 25 kept terms in this collection.
Could anybody know why and how it is solved?
09-07-2017 09:04 AM
If your discovering topics, you need to have at least a small amount of data in order to learn about the relationships within the data and discover topics. In your case, is it possible that your documents are very short, or have common words that might be blocked by the stop list? Also i think rare words that occur in only a few documents are not used. So, is there a way you can make your documents longer or can add a few more ? Make sure you are processing the text and not the pathname as text as well.