import re

def abbreviate(words):
    words = re.split(r'(?:[^a-zA-Z\'])', words)
    acro = ""
    for item in words:
        acro += item[:1]
    return acro.upper()
    return w.group()