Do residential time-of-use tariff policies gain customers’ favor? Insight from a social network-enhanced BERT-BiLSTM analysis
Year: 2025
Type: article
Abstract: Public opinion on social media is a crucial data source for mining public policy needs. Existing studies often assume that social user-generated content is independent and identically distributed, ignoring the interaction between users’ content and their social network relationships, and leading to the misjudgment of public opinion. Therefore, we propose a Social Network Enhanced BERT-BiLSTM (SNE-BERT-BiLSTM) model and apply it to the public’s online opinion mining of China’s time-of-use tariff policy. Integrated with the BERTopic model, this approach was applied to mine public online opinions regarding residential time-of-use tariff policy in China. The results show the following: (1) Compared with the BERT-BiLSTM model without considering the social network, the SNE-BERT-BiLSTM model constructed in this study improved the accuracy of public policy sentiment recognition by 9.19 %. Compared with the baseline model, which considers the network structure, the recognition accuracy was improved by 6.21 %. (2) The Chinese public is generally unsupportive of time-of-use tariff policies, with significant differences in seasonal, geographic, and social network influences. (3) The primary reasons for the lack of public support for time-of-use tariff policies are the additional economic burden imposed on households, the complexity of decision-making processes, misunderstandings of the meaning of the policies, and the dissemination of false information.
Source: Energy Reports
Authors: Shiwei Yu, Mengmeng Jin, Kerui Weng
Institution: China University of Geosciences
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $3,040
Awards: G4717449724, G5290350335, G6933701639
OpenAlex ID: W4416657606