Sentiment Analysis and Research on Consumer Evaluation of Online Shopping Platform Based on Integrated Algorithm
This paper performs sentiment analysis and classification on consumer evaluation data from online shopping platforms.By using Python to realize automatic browser driving and anti-crawler technology,it successfully collects consumer evaluation information of a certain shopping platform.This paper proposes an improved integration algorithm,which uses LSTM,BiGRU and BiLSTM as classifiers,and uses Voting and Bagging methods for integration respectively.The results show that compared with the traditional Bayesian and logistic regression,the LSTM+Bagging integration algorithm improves the accuracy by 5.9％and 6％,respectively,and compared with the LSTM+Voting integration algorithm,the accuracy increases by 0.5 percentage points.In addition,the LSTM+Bagging model outperforms the LSTM+Voting algorithm in terms of stability and robustness.
国家科技期刊平台
NSTL
万方数据
