Does Microsoft azure have a Cognitive services to classify pdf or word files?

Question

I am new to Microsoft cognitive services and have go through the custom vision where we can classify the images that can be classified on the run. do we have some similar product where we can upload a .PDF or word file and it returns the category based on the previous training .

Have got my hands dirty with ML studio of Azure as well but seems it doesn't accept PDF and word file

Answer 1

Vikas - There is no out of box document classifier. But you should be able to build one.

Reference : http://www.sharepointtweaks.com/2018/04/auto-classify-Office365-content-using-azure-machine-learning-studio-part2..html

In this article, basically they are training the model with BBC News Data. They are training using text of the News and Category.

Having said that, here you can probably train with your own dataset.

Also, the model that has been trained & deployed - you will have to pass the data in the same was it was trained. So in the above case - a flow was used to extract the text data from the uploaded file and the extracted text was sent to the trained model.