| 摘 要: 国内电商网站的快速发展促使产生大量的中文商品评论信息。对这些评论进行情感分类有利于获取其中 的有用信息，具有重要的应用意义。目前，情感分类的研究主要基于情感词典或者传统机器学习。这些方法通常需要 人工选取特征，费事费力，分类效果不好。针对这些不足，本文提出一种基于注意力机制的双向LSTM模型，对中文商 品评论进行情感分类。实验结果表明，该模型在中文商品评论二分类任务和三分类任务中均获得了较好的准确率、召回 率、F1值。 |
| 关键词: 中文商品评论；情感分类；注意力机制；双向LSTM |
|
中图分类号: TP391
文献标识码: A
|
|
| Research on the Attention Mechanism-Based Bidirectional LSTM Model for the Sentiment Classification of Chinese Product Reviews |
|
CHENG Lu
|
( Department of Computer Engineer, Taiyuan Institute of Technology, Taiyuan 030008, China)
|
| Abstract: With the rapid development of domestic E-commerce websites,there are lots of Chinese product reviews.The sentiment classification of Chinese product reviews is helpful to obtain useful information,with great application significance. Currently,most sentiment classification studies are based on the sentiment dictionary or traditional machine learning methods.These methods usually need artificial selection of features,with low classification efficiency and effectiveness.In view of all these deficiencies,the paper proposes an attention mechanism-based bidirectional LSTM model for the sentiment classification of Chinese product reviews.The experimental results show that the proposed model has better precision rate,recall rate and F1 score in binary classification tasks and three classification tasks in Chinese product reviews. |
| Keywords: Chinese product reviews;sentiment classification;attention mechanism;bidirectional LSTM |
