Excel False Positives

[Mra885]

I'm using ePO 5.1 with DLP 11.1

Currently, I am doing discovery scans to test policies put in place and I am having a high number of false positives within Excel documents. I was instructed that the proximity classification criteria does not work the same was as it does with text documents. Is this correct? If so what steps can I take besides using advanced patterns and dictionaries? Right now each classification is separated with one focusing on excel and a second one that looks at normal text documents. 

Excel - Advanced Pattern / Dictionary /  File Extension = Only Excel Documents and Other Spreadsheet

Files

For example I am searching for Driver's Licenses - I receive false positives only inside excel documents

What other logic could I add to help remediate this problem?