Smart City Security: Fake News Detection in Consumer Electronics

The rapid and widespread adoption of consumer electronics and technology, such as smartphones, smart speakers, and Internet of Things devices, has fundamentally changed how we access and consume information. This technological revolution has not only enhanced our daily convenience but has also driven the development of smart cities and brought significant advancements in urban living. However, this digital convenience comes with challenges, notably the production of fake news. Through various consumer electronics platforms, false information can be quickly produced and spread, undermining public trust and social order. Although fake news-detecting technology has advanced rapidly benefitting from deep-learning techniques, it often fails to consider feature interactions. To address this issue, we propose an enhanced feature interactions network (EFI-Net) for fake news detection. Specifically, the EFI-Net introduces an efficient additive learning module to enhance feature interaction for language models at different scales. Experiments were conducted using the ARG fake news detection dataset, and the proposed network achieves an accuracy of 88.9% on English fake news and 78.7% on Chinese fake news, which outperforms the state-of-the-art method by a large margin. This work has substantial implications for consumer electronics and technology. Users can benefit from more reliable information filtering and verification by integrating EFI-Net into various consumer electronic platforms.