sentiment.py 文件源码

def getAllCritics(movieList):
    reviews = np.array(map(lambda x: x["critics"], movieList))
    reviews = np.concatenate(reviews)

    tokenizeReview = []

    for review in reviews:
        s = review['review']
        s = RegexpTokenizer(r'\w+').tokenize(s.lower())
        s = map(lambda x: PorterStemmer().stem(x), s)
        s = filter(lambda x: x not in stopwords.words('english'), s)
        tokenizeReview.append((s, 'pos' if review["tomatometer"] == "fresh" else 'neg'))

    return tokenizeReview