The exponential growth of sensitive patient information and diagnostic records in digital healthcare systems has increased the complexity of data protection, while frequent medical data breaches severely compromise system security and reliability. Existing privacy protection techniques often lack robustness and real-time capabilities in high-noise, high-packet-loss, and dynamic network environments, limiting their effectiveness in detecting healthcare data leaks. To address these challenges, we propose a Swarm Intelligence-Based Network Watermarking (SIBW) method for real-time privacy data leakage detection in digital healthcare systems. SIBW integrates fountain codes with outer error correction codes and employs a Multi-Phase Synergistic Swarm Optimization Algorithm (MPSSOA) to dynamically optimize encoding parameters, significantly enhancing the robustness and interference resistance of watermark detection. Additionally, a reliable synchronization sequence and lightweight embedding mechanism are designed to ensure adaptability to complex, dynamic networks. Experimental results demonstrate that SIBW achieves over 90% detection accuracy under high latency jitter and packet loss conditions, surpassing existing methods in both robustness and efficiency. With a compact design of only 3.7 MB, SIBW is particularly suited for rapid deployment in resource-constrained digital healthcare systems.

Automated assessment of open responses in K–12 science education poses significant challenges due to the multimodal nature of student work, which often integrates textual explanations, drawings, and handwritten elements. Traditional evaluation methods that focus solely on textual analysis fail to capture the full breadth of student reasoning and are susceptible to biases such as handwriting neatness or answer length. In this paper, we propose a novel LLM-augmented multimodal evaluation framework that addresses these limitations through a comprehensive, bias-corrected grading system. Our approach leverages LLMs to generate causal knowledge graphs that encapsulate the essential conceptual relationships in student responses, comparing these graphs with those derived automatically from the rubrics and submissions. Experimental results demonstrate that our framework improves grading accuracy and consistency over deep supervised learning and few-shot LLM baselines.

Haohao Zhu, Junyu Lu, Zeyuan Zeng, Zewen Bai, Xiaokun Zhang, Liang Yang, Hongfei Lin. Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers). 2025.

Hateful memes are prevalent on the Internet, raising the urgent need for effective detection.Given their implicit nature, incorporating rationales with background knowledge is crucial for enhancing model understanding.However, existing methods often suffer from limited quality of external rationales and misalignment with original meme information.These challenges hinder model comprehension, leading to reduced accuracy and explainability.To address these challenges, we propose a Multimodal Multi-agent Knowledge Enhanced (M2KE) framework for hateful meme detection.M2KE introduces a multi-agent rationale discovery mechanism to extract high-quality rationales relevant to meme content and an adaptive knowledge interaction mechanism to ensure alignment between original meme information and external rationales.Specifically, multi-agent rationale discovery mechanism improves the reliability of rationales by collaboratively verifying and refining them with multiple agents, supported by large language models (LLMs) due to their extensive knowledge.And adaptive knowledge interaction mechanism uses information entropy to dynamically balance the model's attention between original meme information and external rationales, preventing over-reliance on rationales and enabling a more comprehensive understanding.Experimental results on three datasets demonstrate that M2KE significantly outperforms existing models.Further analysis underscores the importance of effectively integrating accurate rationales to enhance model performance.Disclaimer: Samples in this paper may be considered offensive.

Harmful memes have proliferated on the Chinese Internet, while research on detecting Chinese harmful memes significantly lags behind due to the absence of reliable datasets and effective detectors.To this end, we present the comprehensive detection of Chinese harmful memes.We introduce ToxiCN MM, the first Chinese harmful meme dataset, which consists of 12,000 samples with fine-grained annotations for meme types. Additionally, we propose a baseline detector, Multimodal Knowledge Enhancement (MKE), designed to incorporate contextual information from meme content, thereby enhancing the model's understanding of Chinese memes.In the evaluation phase, we conduct extensive quantitative experiments and qualitative analyses on multiple baselines, including LLMs and our MKE. Experimental results indicate that detecting Chinese harmful memes is challenging for existing models, while demonstrating the effectiveness of MKE.