基于电影评论文本的LSTM情感分析
LSTM sentiment analysis based on documentary review texts
刘晏男 1杨凯 1董小刚1
作者信息
- 1. 长春工业大学 数学与统计学院,吉林 长春 130012
- 折叠
摘要
首先使用CBOW算法对通过 Python 爬取的评论进行了高效且低维的向量表示,然后采用LSTM模型对这些向量进行训练.通过实验对比了 LSTM模型与朴素贝叶斯、决策树、随机森林以及 RNN在预测能力方面的表现,并提供了全面的模型比较.实验结果表明,LSTM模型的准确率更高,具有一定的适用性.
Abstract
In this paper,the CBOW algorithm is first used to make efficient and low-dimensional vector representations of comments crawled by Python,and then these vectors are trained using the LSTM model.This paper also compares the performance of the LSTM model with Naive Bayes,decision trees,random forests,and RNN in terms of predictive ability through experiments.The paper is rigorously structured.A comprehensive model comparison is provided.Experimental results show that the LSTM model has higher accuracy and has certain applicability.
关键词
长短期记忆神经网络模型/Word2vec/评论文本/情感分析Key words
long short-term memory/Word2vec/comment on text/sentiment analysis引用本文复制引用
基金项目
国家自然科学基金项目(11901053)
吉林省自然科学基金项目(20220101038JC)
出版年
2024
NETL
NSTL
万方数据