[Apologies - previously posted this in the wrong thread  :redface:]

In an attempt to organise my life, I bought a ScanSnap scanner and have been scanning (and occasionally shredding) all sorts of stuff - bills, receipts, etc etc.

A faint irritation is the requirement to organise the files post-scan.  I hoped to be able to use some sort of tagging system, but couldn't find anything suitable.  Picasa would have been great, but I am scanning to pdf (for the post-scan search capability) and Picasa only does jpg files.  So I have many folders.

What I would really like is a program which would examine pdf files in a folder and, based on past history, recognise them and suggest a folder (and/or tag) for them.  So rather than scanning and separately filing credit card bills, utility bills, etc, I would just scan and run the program on the unfiled files.  Then I would accept suggestions or create a new folder (thus advancing the learning for the future).

It's very similar to Picasa's face recognition software:  when it detects new photos of people it recognises, it suggests who they are based on past history.  Presumably the suggestion would be based on some sort of jpg recognition algorithm - my AmEx bills all look pretty similar! - but there is also text from the pdf aspect which could be used.

Is this the sort of thing which would work for NANY 2011?  Any interest?

MET