With the popularization of digital technology, the problem of information pollution caused by fake news has become more common. Malicious dissemination of harmful, offensive or illegal content may lead to misleading, misunderstanding and social unrest, affecting social stability and sustainable economic development. With the continuous iteration of artificial intelligence technology, researchers have carried out automatic and intelligent news data mining and analysis based on aspects of information characteristics and realized the effective identification of fake news information. However, the current research lacks the application of multidisciplinary knowledge and research on the interpretability of related methods. This paper focuses on the existing fake news detection technology. The survey includes fake news datasets, research methods for fake news detection, general technical models and multimodal related technical methods. The innovation contribution is to discuss the research progress of fake news detection in communication, linguistics, psychology and other disciplines. At the same time, it classifies and summarizes the explainable fake news detection methods and proposes an explainable human-machine-theory triangle communication system, aiming at establishing a people-centered, sustainable human–machine interaction information dissemination system. Finally, we discuss the promising future research topics of fake news detection technology.

The high level of technological development has enabled fake news to spread faster than real news in cyberspace, leading to significant impacts on the balance and sustainability of current and future social systems. At present, collecting fake news data and using artificial intelligence to detect fake news have an important impact on building a more sustainable and resilient society. Existing methods for detecting fake news have two main limitations: they focus only on the classification of news authenticity, neglecting the semantics between stance information and news authenticity. No cognitive-related information is involved, and there are not enough data on stance classification and news true-false classification for the study. Therefore, we propose a fake news analysis method based on stance information for explainable fake news detection. To make better use of news data, we construct a fake news dataset built on cognitive information. The dataset primarily consists of stance labels, along with true-false labels. We also introduce stance information to further improve news falsity analysis. To better explain the relationship between fake news and stance, we use propensity score matching for causal inference to calculate the correlation between stance information and true-false classification. The experiment result shows that the propensity score matching for causal inference yielded a negative correlation between stance consistency and fake news classification.

Online Public Opinion Reports consolidate news and social media for timely crisis management by governments and enterprises. While large language models (LLMs) enable automated report generation, this specific domain lacks formal task definitions and corresponding benchmarks. To bridge this gap, we define the Automated Online Public Opinion Report Generation (OPOR-Gen) task and construct OPOR-Bench, an event-centric dataset with 463 crisis events across 108 countries (comprising 8.8 K news articles and 185 K tweets). To evaluate report quality, we propose OPOR-Eval, a novel agent-based framework that simulates human expert evaluation. Validation experiments show OPOR-Eval achieves a high Spearman’s correlation (ρ = 0.70) with human judgments, though challenges in temporal reasoning persist. This work establishes an initial foundation for advancing automated public opinion reporting research.

Recent years have witnessed online multiple object tracking (MOT) revealing its enormous potential in numerous fields. However, existing MOT models are not suitable for deployment and application on non-GPU platforms (e.g. CPU-based computers), due to their inefficient inference. In this paper, we offer an automated approach for designing CPU-efficient multi-object tracking architectures, called JDT-NAS. Specifically, we first design several lightweight basic units as candidate operations to construct a heterogeneous search space for multi-object tracking. Then, based on the predefined space, we develop a four-stage hardware-aware search strategy to achieve the automated design of MOT models. Benefiting from our heterogeneous search space and special search strategy, our JDT-NAS not only allows flexible adjustment of the network scale (including depth, width and resolution dimensions), but also better meets the needs of efficient deployment. Finally, on publicly available standard datasets (e.g. MOT17), we apply JDT-NAS to search for multi-object trackers for fast inference on non-GPU computers. Comprehensive experiments show that these searched MOT architectures achieve superior speed and comparable accuracy compared to their state-of-the-art counterparts. This also indicates that our proposed JDT-NAS is a practical and feasible solution for efficient MOT architecture design.

For real-time object detectors, accuracy and efficiency are two important considerations. In this paper, we propose a lightweight anchor-free detector, MAOD, to better balance efficiency and accuracy. Our object detector contains three components: an efficient backbone network (MobileDet), a lightweight feature pyramid structure (L-FPN) and an anchor-free per-pixel prediction method. MobileDet and L-FPN provide more accurate and faster multi-scale feature extraction. Our anchor-free per-pixel prediction method achieves efficient classification and location regression tasks. On the benchmark MS-COCO dataset, MAOD achieves 46.1% AP at the speed of 68 FPS with the input size 512 × 512. When the input size is 800 × 800, MAOD achieves 47.1% AP at the speed of 43 FPS. The fast version of MAOD (320 × 320 input size) can run at 91 FPS with 43.3% AP. Compared with other state-of-the-art object detectors, our detector has similar accuracy while maintaining extremely fast inference speed. MAOD achieves an optimal efficiency-accuracy tradeoff.

Public Opinion Reports are essential tools for crisis management, yet their evaluation remains a critical bottleneck that often delays response actions. Recently, dominant Large Language Model (LLM)-based evaluators often overlook a critical challenge: highly open-ended dimensions such as “innovation” and “feasibility” require synthesizing diverse stakeholder perspectives, as different groups judge these qualities from fundamentally different perspectives. Motivated by this, we propose the Role-based Adaptive Evaluation (RAE) framework. This framework employs an adaptive mechanism leveraging multi-perspective evaluation insights through role-based analysis, and further introduces dynamically generated roles tailored to specific contexts for these dimensions. RAE further incorporates multi-role reasoning aggregation to minimize individual biases and enhance evaluation robustness. Extensive experiments demonstrate that RAE significantly improves alignment with human expert judgments, especially on challenging highly open-ended dimensions.