Blue得分详细计算过程

# 计算BLEU值
def bleu(candidate, references, weights)

# （1）私有函数，计算修正的n元精确率（Modified n-gram Precision）
def _modified_precision(candidate, references, n)

# （2）私有函数，计算BP惩罚因子
def _brevity_penalty(candidate, references)

def _modified_precision(candidate, references, n):
    counts = Counter(ngrams(candidate, n))

    if not counts:
        return 0

    max_counts = {}
    for reference in references:
        reference_counts = Counter(ngrams(reference, n))
        for ngram in counts:
            max_counts[ngram] = max(max_counts.get(ngram, 0), reference_counts[ngram])

    clipped_counts = dict((ngram, min(count, max_counts[ngram])) for ngram, count in counts.items())

    return sum(clipped_counts.values()) / sum(counts.values())

def _brevity_penalty(candidate, references):

    c = len(candidate)
    ref_lens = (len(reference) for reference in references)
    #这里有个知识点是Python中元组是可以比较的，如(0,1)>(1,0)返回False，这里利用元组比较实现了选取参考翻译中长度最接近候选翻译的句子，当最接近的参考翻译有多个时，选取最短的。例如候选翻译长度是10，两个参考翻译长度分别为9和11，则r=9.
    r = min(ref_lens, key=lambda ref_len: (abs(ref_len - c), ref_len))
    print 'r:',r

    if c > r:
        return 1
    else:
        return math.exp(1 - r / c)