Python 情感分析_python情感分析输出积极情感词消极情感词前十名

今天修改了情感分析的程序发现之前有一些不足。这个最简单的实现一个string情感分析的小函数，加载了积极词典，消极词典，程度词典，以及一些反转词等的词典。这里我没有做符号的分析和判断，因为的东西暂时用不到，需要的童鞋可以自己添加。。。

由于这里不能放文件所以...我放到网盘啦，，哇哇哇 链接: http://pan.baidu.com/s/1mhDqfWG  密码: vm38

import jieba
import cPickle as pickle
"""载入情感词典"""
g_pos_dict = pickle.load(open("../sentiment_dic/posdict.pkl", "r"))
g_neg_dict = pickle.load(open("../sentiment_dic/negdict.pkl", "r"))
g_most_dict = pickle.load(open("../sentiment_dic/mostdict.pkl", "r"))
g_more_dict = pickle.load(open("../sentiment_dic/moredict.pkl", "r"))
g_very_dict = pickle.load(open("../sentiment_dic/verydict.pkl", "r"))
g_ish_dict = pickle.load(open("../sentiment_dic/ishdict.pkl", "r"))
g_insufficient_dict = pickle.load(open("../sentiment_dic/insufficentdict.pkl", "r"))
g_inverse_dict = pickle.load(open("../sentiment_dic/inversedict.pkl", "r"))
 
 
def emotion_test(string):
    """情感分析函数"""
    words = list(jieba.cut(string))
    a = 0  # 记录情感词位置
    poscount = 0  # 积极词的分值
    poscount3 = 0  # 积极词最后分值
    negcount = 0
    negcount3 = 0
    for index, word in enumerate(words):
        word = word.encode("utf8")
        if word in g_pos_dict:  # 判断词语是否是积极情感词
            poscount += 1
            c = 0  # 反转词
            for w in words[a:index]:
                w = w.encode("utf8")
                if w in g_most_dict:
                    poscount *= 4.0
                elif w in g_very_dict:
                    poscount *= 3.0
                elif w in g_more_dict:
                    poscount *= 2.0
                elif w in g_ish_dict:
                    poscount /= 2.0
                elif w in g_insufficient_dict:
                    poscount /= 4.0
                elif w in g_inverse_dict:
                    c += 1
            if c % 2 == 1:
                poscount *= -1.0
            poscount3 += poscount
            poscount = 0
            a = index + 1  # 情感词的位置变化
        elif word in g_neg_dict:  # 消极情感
            negcount += 1
            d = 0  # 反转词
            for w in words[a:index]:
                w = w.encode("utf8")
                if w in g_most_dict:
                    negcount *= 4.0
                elif w in g_very_dict:
                    negcount *= 3.0
                elif w in g_more_dict:
                    negcount *= 2.0
                elif w in g_ish_dict:
                    negcount /= 2.0
                elif w in g_insufficient_dict:
                    negcount /= 4.0
                elif w in g_inverse_dict:
                    d += 1
            if d % 2 == 1:
                negcount *= -1.0
            negcount3 += negcount
            negcount = 0
            a = index + 1
    if poscount3 <= 0 and negcount3 <= 0:
        t = poscount3
        poscount3 = -negcount3
        negcount3 = -t
    elif poscount3 <= 0 and negcount3 >= 0:
        negcount3 -= poscount3
        poscount3 = 0
    elif poscount3 >= 0 and negcount3 <= 0:
        poscount3 -= negcount3
        negcount3 = 0
    return poscount3, negcount3
 
 
print emotion_test("我真的非常的烦心")