similarity_score function

def similarity_score(s1, s2):
    largest = []
    for item1 in s1:
        max = 0
        for item2 in s2:
            similarity = wn.wup_similarity(item1, item2)
            print(similarity)
            if similarity:
                if similarity > max:
                    max = similarity
        largest.append(max)
        max = 0
    score = sum(largest) / len(largest)
    return score

I tried to write like this, but when I compute like this,

    synset1 = doc_to_synsets('I like cats')
    synset2 = doc_to_synsets('I like dogs')
    res = similarity_score(synset1, synset2)

the res is 0.9285714285714286, isn't right ,do you have any ideas how to solve this question?