Miao He

In this paper, we introduce two novel, publicly available datasets that capture public opinion on two highly salient and timely topics: the perceived feasibility of artificial general intelligence (AGI) and the Hamas–Israel conflict (HIC). Collected from social media posts on X(formerly Twitter), the datasets were curated and annotated using a structured methodology designed to ensure robustness, consistency, and interpretability. Whereas prior research on public opinion has primarily relied on theoretical models, controlled experiments, or narrowly scoped datasets, such approaches often fail to capture the complexity and dynamism of real-world discourse. In contrast, the datasets presented in this work provide temporally fine-grained, large-scale empirical data that reflect the evolving nature of public sentiment on contemporary global issues. In addition to detailed sentiment annotations, these resources support longitudinal analyses of opinion dynamics, offering a foundation for empirical studies on consensus formation, polarization, and social influence in digital environments. By releasing these datasets, we aim to advance the empirical study of public opinion in the age of social media.

Haihua Xie, Yinzhu Cheng, Yaqing Wang, Miao He, Mingming Sun. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing. 2025.

In this paper, we explore the emerging role of graph neural networks (GNNs) in optimizing routing for next-generation communication networks. Traditional routing protocols, such as OSPF or the Dijkstra algorithm, often fall short in handling the complexity, scalability, and dynamic nature of modern network environments, including unmanned aerial vehicle (UAV), satellite, and 5G networks. By leveraging their ability to model network topologies and learn from complex interdependencies between nodes and links, GNNs offer a promising solution for distributed and scalable routing optimization. This paper provides a comprehensive review of the latest research on GNN-based routing methods, categorizing them into supervised learning for network modeling, supervised learning for routing optimization, and reinforcement learning for dynamic routing tasks. We also present a detailed analysis of existing datasets, tools, and benchmarking practices. Key challenges related to scalability, real-world deployment, explainability, and security are discussed, alongside future research directions that involve federated learning, self-supervised learning, and online learning techniques to further enhance GNN applicability. This study serves as the first comprehensive survey of GNNs for routing optimization, aiming to inspire further research and practical applications in future communication networks.

Satellite-terrestrial integrated networks and space-air-ground integrated networks have been recognized as important enablers of 6G. With more parties involved, resource management in satellite-terrestrial integrated networks becomes much more complex than before, and challenges arise, for example, dynamic network topology, entangled optimization objective, and high algorithmic complexity. Matching games have proven to be effective tools for addressing these challenges. However, there is a lack of an up-to-date summary of existing studies that apply matching games to satellite-terrestrial integrated networks. This study aims to fill this research gap by discussing both two-sided and three-sided matching games in satellite-terrestrial integrated networks, focusing on two research questions: who to match and how to match? Research challenges and opportunities have also been discussed to inspire follow-up studies.

Community-driven communication network re-search is an emerging research paradigm for handling the increasing network research challenges. This paper presents a perspective study for the rise of community-driven communication network research, which fills the research gap in which there is no existing summary. Community-driven research schemes are categorized into three types: data sharing, facility sharing and intelligence sharing, with some representative research projects reviewed and discussed. However, there is still a huge research gap of encouraging more relevant projects with the aim of boosting intelligence sharing for heterogeneous networks. Finally, research challenges and opportunities are highlighted to inspire follow-up studies.

Internet of Things (IoT) is an important infrastructure for supporting vertical applications. However, existing IoT systems are still facing some challenges, e.g., lack of coverage in rural areas and lack of efficient data collection methods. To overcome these challenges, a satellite IoT framework is proposed in this study as a promising solution, and smart agriculture is used as a typical application scenario. A satellite edge computing workflow is further proposed, with computer vision as a case study. In the case study, a lightweight deep learning model named MobileViT is proven effective for aphid detection and infestation severity classification on lemon leaves.