Recent Surge in Public Interest in Transportation: Sentiment Analysis of Baidu Apollo Go Using Weibo Data

Shiqi Wang,Zhouye Zhao,Yuhang Xie,Mingchuan Ma,Zirui Chen,Zeyu Wang,Bohao Su,Wenrui Xu,Tianyi Li
2024-08-19
Abstract:Urban mobility and transportation systems have been profoundly transformed by the advancement of autonomous vehicle technologies. Baidu Apollo Go, a pioneer robotaxi service from the Chinese tech giant Baidu, has recently been widely deployed in major cities like Beijing and Wuhan, sparking increased conversation and offering a glimpse into the future of urban mobility. This study investigates public attitudes towards Apollo Go across China using Sentiment Analysis with a hybrid BERT model on 36,096 Weibo posts from January to July 2024. The analysis shows that 89.56\% of posts related to Apollo Go are clustered in July. From January to July, public sentiment was mostly positive, but negative comments began to rise after it became a hot topic on July 21. Spatial analysis indicates a strong correlation between provinces with high discussion intensity and those where Apollo Go operates. Initially, Hubei and Guangdong dominated online posting volume, but by July, Guangdong, Beijing, and international regions had overtaken Hubei. Attitudes varied significantly among provinces, with Xinjiang and Qinghai showing optimism and Tibet and Gansu expressing concerns about the impact on traditional taxi services. Sentiment analysis revealed that positive comments focused on technology applications and personal experiences, while negative comments centered on job displacement and safety concerns. In summary, this study highlights the divergence in public perceptions of autonomous ride-hailing services, providing valuable insights for planners, policymakers, and service providers. The model is published on Hugging Face at <a class="link-external link-https" href="https://huggingface.co/wsqstar/bert-finetuned-weibo-luobokuaipao" rel="external noopener nofollow">this https URL</a> and the repository on GitHub at <a class="link-external link-https" href="https://github.com/GIStudio/trb2024" rel="external noopener nofollow">this https URL</a>.
Social and Information Networks
What problem does this paper attempt to address?
The main problem that this paper attempts to solve is to understand the public's attitude and reaction towards Baidu Apollo Go, an autonomous taxi service. Specifically, through analyzing Weibo data, the research explored the following aspects: 1. **How to use natural language data to better understand new mobility services?** - The research used natural language processing (NLP) techniques to conduct sentiment analysis on text data on Weibo in order to understand the public's views on autonomous taxis. 2. **What are the public's emotional reactions to the sudden emergence of autonomous taxis?** - Through sentiment analysis, the research revealed the changes in the public's positive, neutral, and negative attitudes towards Apollo Go in different time periods. 3. **What suggestions and insights can be provided for the public and transportation agencies?** - The research results provide valuable references for policymakers and service providers, helping them better understand and respond to the public's attitude towards autonomous taxis. ### Specific Problems and Their Solutions #### 1. Data Collection and Processing - **Data Source**: The research used 36,096 Weibo posts and comments with the "Baidu Apollo Go" tag from January to July 2024. - **Data Cleaning**: Irrelevant content and content with no clear attitude published by official media were removed, and long - text was manually labeled. - **Data Classification**: The data was divided into training sets and validation sets, and classified into positive, neutral, and negative categories according to sentiment. #### 2. Sentiment Analysis Method - **Model Selection**: A pre - trained Chinese BERT model (`bert - base - chinese`) was selected and fine - tuned to adapt to Weibo data. - **Hyperparameter Settings**: - Learning Rate: \(1\times10^{-5}\) - Batch Size: 8 - Number of Training Rounds: 5 - Optimizer: Adam (\(\beta_1 = 0.9\), \(\beta_2 = 0.999\), \(\epsilon = 1\times10^{-8}\)) - Learning Rate Scheduler: Linear Decay #### 3. Result Analysis - **Time - Feature Analysis**: - It was found that 89.56% of relevant posts were concentrated in July, which was in line with the time when Apollo Go officially started commercial operations in Wuhan and became a hot topic. - From January to July, the public's attitude was generally positive, but as the popularity of the discussion increased, the proportion of negative comments also began to rise. - **Space - Feature Analysis**: - In the first six months, Hubei and Guangdong were the most actively discussed provinces, and in July, the discussion volume in Beijing and Guangdong increased significantly. - There are differences in the public's attitude in different regions. The eastern coastal provinces such as Beijing, Guangdong, and Shandong have a higher discussion intensity, mainly focusing on employment issues and the possibility of technology replacing human jobs. ### Main Conclusions - The public's attitude towards autonomous taxis is diverse, with both support for its technological innovation and convenience, and concerns about its safety and employment impact. - The research results provide valuable references for policymakers and service providers, helping them better consider the public's opinions and needs when promoting autonomous driving technology. Through these analyses, the research not only reveals the changes in the public's attitude towards autonomous taxis, but also provides important data support for future research and practice.