检索结果-内蒙古大学图书馆

31st International Conference on Computational Linguistics, COLING 2025

作者： Wang, Shiyu Zhou, Gang Lu, Jicang Chen, Jing Huang, Ningbo State Key Laboratory of Mathematical Engineering and Advanced Computing China

ISBN: (纸本)9798891761964

Nowadays, research of Text Classification (TC) based on graph neural networks (GNNs) is on the rise. Both inductive methods and transductive methods have made significant progress. For transductive methods, the semantic interaction between texts plays a crucial role in the learning of effective text representations. However, it is difficult to perform inductive learning while modeling interactions between texts on the graph. To give a universal solution, we propose the graph neural network based on pre-trained semantic interaction called PaSIG. Firstly, we construct a text-word heterogeneity graph and design an asymmetric structure to ensure one-way message passing from words to the test texts. Meanwhile, we use the context representation capability of the pre-trained language model to construct node features that contain classification semantic information. Afterward, we explore the adaptative aggregation methods with a gated fusion mechanism. Extensive experiments on five datasets have shown the effectiveness of PaSIG, with the accuracy exceeding the baseline by 2.7% on average. While achieving state-of-the-art performance, we have also taken measures of subgraph sampling and intermediate state preservation to achieve fast inference. © 2025 Association for Computational Linguistics.

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Research on Intrusion Tolerance System Based on Behavioral Baseline and Defense-in-Depth Architecture

Research on Intrusion Tolerance System Based on Behavioral B...

引用

2024 International Conference on Mechatronics and Intelligent Control, ICMIC 2024

作者： Meng, Qi Wang, Xuemeng Yang, Kai Meng, Long Wang, Jiancheng Fei, Jinlong State Key Laboratory of Mathematical Engineering and Advanced Computing PLA Strategic Support Force Information Engineering University Zhengzhou China

ISBN: (数字)9781510686847

ISBN: (纸本)9781510686830

Tolerance technology is the third-generation network security technology commonly used in the world. It is derived from the category of information survival and endogenous security technology. A scholar from Carnegie Mellon University gave this survival technology a definition: the so-called "invasion survival technology".It is the ability of the system to perform its own tasks within a limited time when external attacks, failures and accidents have occurred. It assumes that we cannot detect intrusions to the system completely and correctly. When external intrusions or failures occur suddenly, we can use tolerance technology to solve the problem of survival of the system to ensure the confidentiality, integrity, and integrity of the information system. reliability and non-repudiation. In the current Internet security incidents, countless experiences and lessons tell us that it is not enough to rely on blocking and defensing for network *** on the existing network defense-in-depth architecture and the common behavior and parameter baselines of current emergency response systems, this paper designs an intrusion-tolerant system based on the behavioral baseline and network-in-depth architecture, which realizes intrusion tolerance and malicious behavior blocking under high confrontation intensity. To adapt to the maximum recovery and necessary tolerance under the attack technology of international APT organizations, and to protect the integrity and availability of the system to the greatest extent. © 2025 SPIE.

关键词： Network intrusion

来源：评论

学校读者我要写书评

暂无评论

Social Media User Geolocation Based on Large Language Models 2nd

Social Media User Geolocation Based on Large Language Model...

引用

2nd International Conference on Data Security and Privacy Protection, DSPP 2024

作者： Zhang, Meng Luo, Xiangyang Huang, Ningbo Key Laboratory of Cyberspace Situation Awareness of Henan Province Henan Zhengzhou China State Key Laboratory of Mathematical Engineering and Advanced Computing Henan Zhengzhou China

ISBN: (纸本)9789819785391

In the digital age, social media platforms have amassed a wealth of user-generated content, which contains valuable geographic information. However, the irregularities and noise in user-generated text, have led to suboptimal performance in traditional text-based user geolocation methods. We propose a unsupervised framework for user geolocation based on Large Language Models (LLMs), which utilizes the LLMs’ powerful text processing abilities to geolocate users based on user-generated text unsupervisedly. Firstly, preprocess the text using regularization rules and LLM to denoise and normalize user-generated text, thus enhancing data quality. Subsequently, appropriate prompts are designed to guide the knowledgeable LLM in understanding the user text’s geolocating mechanism, thereby profiling users. To refine user geolocation accuracy, five independent positioning iterations are conducted, with the most frequent occurrence identified as the final user location. Through a series of experiments, we have demonstrated the potential of utilizing large language models for processing noisy text and the effectiveness of geolocating users in an unsupervised setting. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Large language model Social media Unsupervised user geolocation User-generated text

来源：评论

学校读者我要写书评

暂无评论

Mitigating Language Confusion through Inference-time Intervention 31

Mitigating Language Confusion through Inference-time Interve...

引用

31st International Conference on Computational Linguistics, COLING 2025

作者： Xie, Yunfan Zou, Lixin Luo, Dan Tang, Min Li, Chenliang Dong, Liming Luo, Xiangyang Wuhan University Wuhan China Lehigh University Bethlehem United States Monash University Melbourne Australia National Defense University Beijing China State Key Lab of Mathematical Engineering and Advanced Computing Zhengzhou China

ISBN: (纸本)9798891761964

Although large language models (LLMs) trained on extensive multilingual corpora exhibit impressive language transfer, they often fail to respond in the user's desired language due to corpus imbalances, an embarrassingly simple problem known as the language confusion. However, existing solutions like in-context learning and supervised fine-tuning (SFT) have drawbacks: in-context learning consumes context window space, diminishing attention as text lengthens, while SFT requires extensive, labor-intensive data collection. To overcome these limitations, we propose the language-sensitive intervention (LSI), a novel, lightweight, and label-free approach. Specifically, we analyze language confusion from a causal perspective, revealing that the training corpus's language distribution acts as a confounder, disadvantaging languages that are underrepresented in the dataset. Then, we identify a language-sensitive dimension in the LLM's residual stream, i.e., the language vector, which allows us to estimate the average causal effect of prompts on this dimension. During inference, we directly intervene on the language vector to generate responses in the desired *** further advance research on this issue, we introduce a new benchmark that detects language confusion and assesses content quality. Experimental results demonstrate that our method effectively mitigates language confusion without additional complex mechanisms. Our code is available at https://***/SoseloX/LSI. © 2025 Association for Computational Linguistics.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

Stance Detection for Social Text: Inference-Enhanced Multi-Task Learning with Machine-Annotated Supervision

Stance Detection for Social Text: Inference-Enhanced Multi-T...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Qiankun Pi Jicang Lu Yepeng Sun Qinlong Fan Xukun Zhou Shouxin Shang State Key Laboratory of Mathematical Engineering and Advanced Computing Zhengzhou China School of Information Renmin University of China Beijing China

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

Stance detection, a natural language processing technique, captures user attitudes on controversial social media topics. However, the semantic ambiguity of social texts makes accurate stance determination challenging. Existing annotated data is often domain-specific, resulting in poor model generalization for unseen targets and cross-domain scenarios. To tackle this challenge, we introduce inference tasks related to stance detection as auxiliary tasks and use a multitask learning approach to improve the model’s understanding of textual semantics. Meanwhile, to alleviate the scarcity of annotated data, we use argumentation corpus with abundant resources as the source domain to train the basic stance detection model. We integrate this model with a large-scale framework to facilitate weakly supervised learning for machine annotation of unlabeled social texts, thereby boosting the model’s adaptability across different targets and domains. Experimental results on four Twitter datasets demonstrate that our method can significantly improve the model’s stance discriminative ability and generalization performance.

关键词： Adaptation models Weak supervision Social networking (online) Annotations Semantics Blogs Transfer learning Signal processing Data models Speech processing

来源：评论

学校读者我要写书评

暂无评论

Stance Detection for Social Text: Inference-Enhanced Multi-Task Learning with Machine-Annotated Supervision

Stance Detection for Social Text: Inference-Enhanced Multi-T...

引用

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Pi, Qiankun Lu, Jicang Sun, Yepeng Fan, Qinlong Zhou, Xukun Shang, Shouxin State Key Laboratory of Mathematical Engineering and Advanced Computing Henan Zhengzhou China School of Information Renmin University of China Beijing China

ISBN: (纸本)9798350368741

Stance detection, a natural language processing technique, captures user attitudes on controversial social media topics. However, the semantic ambiguity of social texts makes accurate stance determination challenging. Existing annotated data is often domain-specific, resulting in poor model generalization for unseen targets and cross-domain scenarios. To tackle this challenge, we introduce inference tasks related to stance detection as auxiliary tasks and use a multitask learning approach to improve the model's understanding of textual semantics. Meanwhile, to alleviate the scarcity of annotated data, we use argumentation corpus with abundant resources as the source domain to train the basic stance detection model. We integrate this model with a large-scale framework to facilitate weakly supervised learning for machine annotation of unlabeled social texts, thereby boosting the model's adaptability across different targets and domains. Experimental results on four Twitter datasets demonstrate that our method can significantly improve the model's stance discriminative ability and generalization performance. © 2025 IEEE.

关键词： argumentation corpus machine annotation multitask learning stance detection

来源：评论

学校读者我要写书评

暂无评论

On the Cycle Structure of Some Nonlinear Feedback Shift Registers

引用

Chinese Journal of Electronics 2025年第4期23卷 801-804页

作者： Zhongxiao Wang Hong Xu Wenfeng Qi State Key Laboratory of Mathematical Engineering and Advanced Computing Zhengzhou Information Science and Technology Institute Zhengzhou China Department of Computer Science and Engineering Shanghai Jiao Tong University Shanghai China

In this paper, we partly determine the cycle structure of two types of Nonlinear feedback shift registers (NFSRs). Based on these results, the cycle structure of a class of NFSRs with symmetric feedback functions can be completely characterized. Furthermore, an alternative proof of Kjeldsen's results is presented. Compared with the original proof based on abstract algebra theory, ours is straightforward and easy to understand.

关键词： Abstract algebra Shift registers Cryptography

来源：评论

学校读者我要写书评

暂无评论

Minimizing transformer inference overhead using controlling element on Shenwei AI accelerator

引用

Frontiers of Information Technology & Electronic engineering 2025年第4期26卷 605-622页

作者： Zhao, Yulong Wu, Chunzhi Wang, Yizhuo Zhang, Lufei Zhang, Yaguang Shen, Wenyuan Fan, Hao Fang, Hankang Qin, Yi Liu, Xin State Key Laboratory of Mathematical Engineering and Advanced Computing Wuxi China School of Non-Commissioned Officer Space Engineering University Beijing China National Supercomputing Center in Wuxi Wuxi China Zhejiang Lab Hangzhou China National Research Centre of Parallel Computer Engineering and Technology Beijing China

Transformer models have become a cornerstone of various natural language processing (NLP) tasks. However, the substantial computational overhead during the inference remains a significant challenge, limiting their deployment in practical applications. In this study, we address this challenge by minimizing the inference overhead in transformer models using the controlling element on artificial intelligence (AI) accelerators. Our work is anchored by four key contributions. First, we conduct a comprehensive analysis of the overhead composition within the transformer inference process, identifying the primary bottlenecks. Second, we leverage the management processing element (MPE) of the Shenwei AI (SWAI) accelerator, implementing a three-tier scheduling framework that significantly reduces the number of host-device launches to approximately 1/10 000 of the original PyTorch-GPU setup. Third, we introduce a zero-copy memory management technique using segment-page fusion, which significantly reduces memory access latency and improves overall inference efficiency. Finally, we develop a fast model loading method that eliminates redundant computations during model verification and initialization, reducing the total loading time for large models from 22 128.31 ms to 1041.72 ms. Our contributions significantly enhance the optimization of transformer models, enabling more efficient and expedited inference processes on AI accelerators.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Robust Watermarking Based on Multi-layer Watermark Feature Fusion

引用

IEEE Transactions on Multimedia 2025年

作者： Wu, Shaowu Lu, Wei Luo, Xiangyang Sun Yat-Sen University School of Computer Science and Engineering Ministry of Education Key Laboratory of Information Technology Guangdong Province Key Laboratory of Information Security Technology Guangzhou510006 China State Key Laboratory of Mathematical Engineering and Advanced Computing Henan450002 China

The purpose of robust image watermarking is to embed a watermark into a carrier image in an invisible form and extract the watermark successfully even under noise interference conditions to achieve copyright confirmation and traceability. Although watermarking methods based on deep learning can improve the robustness by adding a noise simulation layer, few theoretical analyses of the codec structure have been conducted. Theoretical explainability is the theoretical basis for developing a network architecture, which plays a guiding role in network development. On the basis of the interpretability of convolutional networks, this paper analyzes the mathematical process of embedding and extracting watermarks in codecs and proposes a novel watermarking framework based on multi-layer watermark feature fusion. Specifically, the encoder can be a convolutional network structure of arbitrary depth, whereas the decoder needs only to adopt its corresponding deconvolution structure. To improve the quality and robustness of the generated watermarked image, the watermark is associated with an arbitrary layer feature space in the decoder. In the decoder, the network quickly converges to each original encoding feature space through the deconvolution structure, thus decoupling the watermark features. Finally, the watermark is extracted via the automatic fusion of multi-layer watermark features. The experimental results show that the proposed method is suitable for few-shot learning, and its invisibility, robustness and generalization performance on multiple datasets are significantly better than those of other advanced methods. © 2025 IEEE.

关键词： Decoding

来源：评论

学校读者我要写书评

暂无评论

Pixel+ and Pixel++: compact and efficient forward-secure multi-signatures for PoS blockchain consensus 24

Pixel+ and Pixel++: compact and efficient forward-secure mul...

引用

Proceedings of the 33rd USENIX Conference on Security Symposium

作者： Jianghong Wei Guohua Tian Ding Wang Fuchun Guo Willy Susilo Xiaofeng Chen State Key Laboratory of Integrated Service Networks (ISN) Xidian University and State Key Laboratory of Mathematical Engineering and Advanced Computing State Key Laboratory of Integrated Service Networks (ISN) Xidian University College of Cyber Science Nankai University School of Computing and Information Technology University of Wollongong

ISBN: (纸本)9781939133441

Multi-signature schemes have attracted considerable attention in recent years due to their popular applications in PoS blockchains. However, the use of general multi-signature schemes poses a critical threat to the security of PoS blockchains once signing keys get corrupted. That is, after an adversary obtains enough signing keys, it can break the immutable nature of PoS blockchains by forking the chain and modifying the history from some point in the past. Forward-secure multi-signature (FS-MS) schemes can overcome this issue by periodically updating signing keys. The only FS-MS construction currently available is Drijvers et al's Pixel, which builds on pairing groups and only achieves forward security at the time period *** this work, we present new FS-MS constructions that either are free from pairing or capture forward security at the individual message level (i.e., fine-grained forward security). Our first construction Pixel+ works for a maximum number of time periods T. Pixel+ signatures consist of only one group element, and can be verified using two exponentiations. It is the first FS-MS from RSA assumption, and has 3.5x and 22.8x faster signing and verification than Pixel, respectively. Our second FS-MS construction Pixel++ is a pairing-based one. It immediately revokes the signing key's capacity of re-signing the message after creating a signature on this message, rather than at the end of the current time period. Thus, it provides more practical forward security than Pixel. On the other hand, Pixel++ is almost as efficient as Pixel in terms of signing and verification. Both Pixel+ and Pixel++ allow for non-interactive aggregation of signatures from independent signers and are proven to be secure in the random oracle model. In addition, they also support the aggregation of public keys, significantly reducing the storage overhead on PoS *** demonstrate how to integrate Pixel+ and Pixel++ into PoS blockchains. As a proof-of-concept, we provid

关键词：

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：