ryfan-rs/chinesesentimentanalysis

【基于电影评论数据的中文情感分析】Chinese Sentiment Analysis based on ML(Machine Learning) and DL(Deep Learning) algorithms, such as Naive-Bayes, SVM, Decision-Tree, KNN, MLP, CNN, RNN(LSTM).

Sentiment Analysis Implementations

Chinese Sentiment Analysis based on ML(Machine Learning) and DL(Deep Learning) algorithms, such as Naive-Bayes, SVM, Decision-Tree, KNN, MLP, CNN, RNN(LSTM).

Requirements

All code in this project is implemented in Python3.6+.
And all the essential packages are listed in requirements.txt, you can install them by pip install -r requirements.txt -i https://pypi.douban.com/simple/
Anaconda or virtualenv + virtualenvwrapper are strongly recommended to manage your Python environments.

Data Preparation

1.数据集
使用电影评论数据作为训练数据集
训练数据集20000(正向10000, 负向10000)
测试数据集6000条(正向3000, 负向3000)

2.数据处理
1.使用jieba进行分词(去除了标点符号，停用词好像没有去除?)
2.使用预训练的词向量模型，对句子进行向量化

训练与对比(准确率)

Algorithm	Accuracy
Naive-Bayes	{avg: 73.72%, fasttext: 74.32%, matrix: %}
Decision-Tree	0.6907
KNN	0.7909
SVM	0.8303
MLP	0.8359
CNN	0.8376
LSTM	0.8505