classifier.py 文件源码

def ngram_baseline(text):
    ngs = ngrams(text, 2)
    cnt = 0
    """
    for t in ngs:
        print(t, )
        cnt = cnt + 1
        if (cnt > 1000):
            break
    """
    refine = []
    for (first, second) in ngs:
        if (second[1] == 1):
            #print(first[0], second[0], zip(first[0], second[0]))
            #tmp = (first[0], second[0])
            #print(tmp)
            #break
            refine.append((first[0], second[0]))
    cnt = 0
    """
    for t in refine:
        print(t)
        cnt = cnt + 1
        if (cnt > 1000):
            break
    #print(ngs)
    """
    cfdist = nltk.ConditionalFreqDist(refine)
    return cfdist