You could do it indirectly by looking at where the miRNA binds, and applying some method to determine the function of the miRNA target. You can identify predicted targets using resources such as http://microrna.org and Targetscan.

So essentially you would be mapping your set of miRNAs to a set of mRNAs, which you would then map to a list of functions and/or pathways. I am less sure about the last step, maybe Gene ontology (GO) enrichment analysis?

There is this database that has published information on miRNAs called miRBase: http://mirbase.org This can be helpful if those miRNA found are in it.

However I'm no certain if this is still being maintained. But has a lot of information still.