Multilingual User Perceptions Analysis from Twitter using Zero Shot Learning for Border Control Technologies

Online social networks (OSNs) such as Twitter, Facebook, Instagram, and Reddit have transformed communication by enabling users to share opinions and perceptions on various topics. The vast amount of user-generated content on these platforms poses significant challenges for manual analysis. Advances in artificial intelligence, particularly transformer-based models like BERT and GPT, have improved the processing of multilingual data for tasks such as text classification, sentiment analysis, and emotion analysis. However, these models often require extensive task-specific training and high-quality labeled data, which is impractical for multilingual contexts. This study addresses these limitations by utilizing zero-shot learning (ZSL) with transformer-based models, which eliminate the need for task-specific training and can classify new data into unseen classes without manual annotation. The use case for this study is border control technologies (BCTs), a hot topic following the European Union (EU) commission’s ”Smart Borders Package” aimed at improving border crossing points’ efficiency and security. We introduce a novel ”user perception extraction architecture” to analyze multilingual perceptions of BCTs from Twitter. As there is no existing multilingual Twitter dataset for this purpose, we compiled a dataset of 90,789 multilingual tweets related to BCTs from 2008 to 2022. This study contributes to the research domain of user perception analysis from OSNs and opens new directions for understanding and improving global public perceptions of BCTs and other technologies or domains.