eliascai/sentiment

CCF大数据比赛，基于主题的文本情感分析

sentiment's Issues

指出几点需要修改的地方(由于第三方库版本更迭)

• 第23行 (第三方库导入)

# from keras.preprocessing.sequence import pad_sequences
from tensorflow.keras.preprocessing.sequence import pad_sequences

• 122行 (由于tensorflow版本更迭)

def getPredWord(model, word_index, x, y):
    # 获取评论内容的预测关键词
    # pos_pred = model.predict_classes(x) # 由于tensorflow新的版本将predict_classes删除了, 替换为下面这两行代码即可解决
    predict_x = model.predict(x)
    pos_pred = np.argmax(predict_x, axis=1) 

非常感谢作者的开源, 代码跑得很快, 再次感谢~

I face to a issuse!

需要的包已经更新，但是遇到这个错误～～～～，不知道作者能否跑通～～～。
Using TensorFlow backend.
('len of contents :', 20000)
('len of words : ', 20000)
关键词 / 被覆盖的关键词
Building prefix dict from the default dictionary ...
Loading model from cache /tmp/jieba.cache
Loading model cost 0.210 seconds.
Prefix dict has been built succesfully.
Traceback (most recent call last):
File "sentiment_words.py", line 56, in
contents, words = calWordCover(contents, words, 15)
File "sentiment_words.py", line 50, in calWordCover
contents[i] = ' '.join(list(tags))
TypeError: list indices must be integers, not unicode