Sentiment analysis of low-carbon travel APP user comments based on deep learning

Xing Min Lin, Chun Heng Ho, Lu Ting Xia, Ruo Yi Zhao

Research output: Contribution to journalArticlepeer-review

Abstract

Online car-hailing has become an important travel mode in today's society, and it is also an important part of the development of smart transportation. Optimizing online car-hailing consumer behavior preferences can promote low-carbon travel, thereby saving energy and reducing pollution. The use of text sentiment analysis methods is an important analytical approach for online ride-hailing APP review data. Traditional machine learning methods and artificially labeled sentiment dictionary methods rely too much on manual operations when performing sentiment analysis. This paper aims to reduce the dependence of sentiment analysis methods on humans and develop more efficient sentiment analysis methods. This study uses an experimental model based on the combination of long and short-term memory neural network (Bi LSTM) and attention mechanism (Attention). Our results verify that the emotion classification method combining the two has higher F1 value, average macro recall rate and average accuracy rate. It proves that the method of combining Attention and Bi LSTM is more accurate and efficient than traditional machine learning techniques.

Original languageEnglish
Article number101014
JournalSustainable Energy Technologies and Assessments
Volume44
DOIs
Publication statusPublished - 2021 Apr

All Science Journal Classification (ASJC) codes

  • Renewable Energy, Sustainability and the Environment
  • Energy Engineering and Power Technology

Fingerprint Dive into the research topics of 'Sentiment analysis of low-carbon travel APP user comments based on deep learning'. Together they form a unique fingerprint.

Cite this