import re

def count_words(text):
    words = re.findall(r"\b[\w]+(?:'[\w]+)?\b", text.lower(), re.UNICODE)
    result = {}

    for word in words:
        result[word] = result.get(word, 0) + 1

    return result