检索结果-内蒙古大学图书馆

LLM4TAP: LLM-Enhanced TAP Rule Recommendation

IEEE Internet of Things Journal 2025年第10期12卷 13157-13169页

作者： Wu, Gang Hu, Liang Hu, Yuxiao Xiong, Xingbo Wang, Feng Jilin University College of Computer Science and Technology Changchun130012 China Changchun University College of Computer Science and Technology Changchun130022 China

Trigger-action programming (TAP) is an Internet of Things (IoT) paradigm that enables nonprofessional end-users to automate smart devices by formulating rules, such as "IF you leave home, THEN turn off lights." As the number of possible rules increases, manually browsing these rules becomes increasingly time-consuming for users. Recently, graph-based recommendation systems have shown promise in automatically suggesting rules, yet they face two issues. First, these studies struggle to identify and differentiate users’ demands (e.g., turning off lights) and intentions (e.g., energy saving). Second, they overlook the issue of sparse user-rule interactions. In this article, we propose LLM4TAP, a large language model (LLM) enhanced TAP rule recommendation framework, to address these issues. Prior to LLM4TAP, a user-rule graph is constructed to represent the interactions between users and rules. Within LLM4TAP, singular value decomposition is first employed to generate an augmented graph, strengthening global structural relationships between users and rules. Next, the reasoning capabilities of LLMs are utilized to infer users’ demands and intentions from the textual descriptions of rules and user-rule interactions, producing representations of these inferred demands and intentions. Finally, a dual representation alignment method is introduced, integrating user demands and intentions derived from LLMs with the global structural information from the augmentation graph within a contrastive learning framework to enhance representation performance. Extensive experiments demonstrate the effectiveness of LLM4TAP, achieving the maximum improvements of 8.96% and 4.72% over the strongest compared methods on the IFTTT and Wyze datasets, respectively. © 2014 IEEE.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Fluorinated poly(p-triphenylene isatin)anion exchange membranes based on hydrophilic hydroxyl side chain modulation for fuel cells

引用

Journal of Energy Chemistry 2025年第1期100卷 702-709页

作者： Yiman Gu Yanchao Zhang Zhanyu Li Yijia Lei Baozeng Sun Xiaoyu Yu Zhe Wang College of Chemistry and Life Science Changchun University of TechnologyChangchun 130012JilinChina College of Chemical Engineering Changchun University of TechnologyChangchun 130012JilinChina School of Materials Science and Engineering Changchun University of TechnologyChangchun 130012JilinChina Key Laboratory of Advanced Functional Polymer Membrane Materials of Jilin Province Changchun 130012JilinChina

The development of alkaline fuel cells is moving forward at an accelerated pace,and the application of ether-free bonded polymers to anion exchange membranes(AEMs)has been widely ***,the question of the“trade-off”between AEM ionic conductivity and dimensional stability remains *** strategy of inducing microphase separation to improve the performance of AEM has attracted much attention recently,but the design of optimal molecular structures is still being ***,this work introduced different ratios of 3-bromo-1,1,1-trifluoroacetone(x=40,50,and 60)into the main chain of poly(p-terphenylene isatin).Because fluorinated groups have excellent hydrophobicity,hydrophilic hydroxyl-containing side chains are introduced to jointly adjust the formation of phase separation *** results show that PTI-PTF_(50)-NOH AEM with the appropriate fluorinated group ratio has the best ionic conductivity and alkali stability under the combined effect of *** has an ionic conductivity of 133.83 mS cm^(-1)at 80°*** addition,the OH-conductivity remains at 89%of the initial value at 80°C and 3 M KOH for 1056 h of *** cell polarization curve based on PTI-PTF_(50)-NOH shows a power density of 734.76 mW cm^(-2)at a current density of 1807.7 mA cm^(-2).

关键词： Anion exchange membranes Poly(p-terphenylene isatin) Hydrophilic-hydrophobic phase separation Fluorinated group

来源：评论

学校读者我要写书评

暂无评论

ConvGRU: A Lightweight Intrusion Detection System for Vehicle Networks Based on Shallow CNN and GRU

引用

IEEE Access 2025年 13卷 73297-73318页

作者： Wang, Shaoqiang Cheng, Jiahui Wang, Yizhe Li, Shutong Kang, Lei Dai, Yin-Fei Changchun University School of Computer Science and Technology Changchun130022 China Jilin University College of Computer Science and Technology Changchun130012 China

The rapid proliferation of connected vehicles has significantly expanded the attack surface of the Internet of Vehicles (IoV), introducing severe security risks. In such resource-constrained environments, developing lightweight solutions is crucial to ensuring real-time detection and efficient deployment. To ad-dress these challenges, this study proposes ConvGRU, a lightweight vehicular network intrusion detection model that integrates a shallow Convolutional Neural Network (CNN) with a Gated Recurrent Unit (GRU). By employing optimizations such as small convolutional kernels and depthwise separable convolutions, the model significantly reduces the number of parameters and computational overhead, making it well-suited for resource-limited IoV environments. The shallow CNN effectively captures spatial features, while the GRU extracts temporal dependencies, enhancing the model's generalization ability. ConvGRU achieves an accu-racy, precision, recall, and F1-score exceeding 0.99 on the HCRL-Car-hacking, OTIDS, and CICIDS-2018 datasets, with only 112.55K parameters and a memory footprint of merely 0.43 MB. Experimental results demonstrate that this intrusion detection solution substantially improves malicious traffic detection accuracy while ensuring efficient operation in resource-constrained vehicular environments. © 2013 IEEE.

关键词： Network intrusion

来源：评论

学校读者我要写书评

暂无评论

PBCounter:weighted model counting on pseudo-boolean formulas

引用

Frontiers of computer science 2025年第3期19卷 55-63页

作者： Yong LAI Zhenghang XU Minghao YIN College of Computer Science and Technology Jilin UniversityChangchun 130012China Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education Jilin UniversityChangchun 130012China School of Computer Science and Information Technology Northeast Normal UniversityChangchun 130017China

In Weighted Model Counting(WMC),we assign weights to literals and compute the sum of the weights of the models of a given propositional formula where the weight of an assignment is the product of the weights of its *** current WMC solvers work on Conjunctive Normal Form(CNF)***,CNF is not a natural representation for human-being in many *** by the stronger expressive power of Pseudo-Boolean(PB)formulas than CNF,we propose to perform WMC on PB *** on a recent dynamic programming algorithm framework called ADDMC for WMC,we implement a weighted PB counting tool *** compare PBCounter with the state-of-the-art weighted model counters SharpSAT-TD,ExactMC,D4,and ADDMC,where the latter tools work on CNF with encoding methods that convert PB constraints into a CNF *** experiments on three domains of benchmarks show that PBCounter is superior to the model counters on CNF formulas.

关键词： weighted model counting pseudo-boolean constraint algebraic decision diagrams preprocessing techniques

来源：评论

学校读者我要写书评

暂无评论

Photoelectron momentum distributions by an inhomogeneous polarization-gated pulse

引用

Optics Express 2025年第6期33卷 14419-14430页

作者： Sun, Yue Yan, Peng Sun, Qi-Hang School of Materials Science and Engineering Changchun University of Technology Changchun130012 China Changchun University of Technology Key Laboratory of Advanced Structural Materials Ministry of Education Changchun130012 China College of Electronic Science and Engineering Jilin University Changchun130012 China

We theoretically investigate the photoelectron momentum distributions (PMDs) of the argon atom by using an inhomogeneous polarization-gated pulse. The numerical results show that the symmetry of the PMDs is disrupted and a vertical interference stripe appears as the inhomogeneity parameters increase. We illustrate that the interference mechanisms of these vertical interference stripes are different from traditional interference mechanisms, which are interference between direct electrons ionized in the same quarter-cycle. In addition, by analyzing the electron trajectories, we find that the breakdown of symmetry is due to the strong suppression of the positive deviation of electrons along the x-axis by the inhomogeneous PG pulse. © 2025 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement.

关键词： Photoelectrons

来源：评论

学校读者我要写书评

暂无评论

Bi-ZFO/BMO-Vo Z-scheme heterojunction photocatalysis-PMS bidirectionally enhanced coupling system for environmental remediation

引用

材料科学技术（英文版） 2025年第5期208卷 164-175页

作者： Zhaoxin Lin Jing Shao JianWei Zhu Dandan Wang School of Chemistry and Environmental Engineering Changchun University of Science and TechnologyChangchun 130022China Zhongshan Institute of Changchun University of Science and Technology Zhongshan 528437China Zhongshan Institute of Changchun University of Science and Technology Zhongshan 528437China College of Engineering Jilin Normal UniversitySiping 136000China Key Laboratory of Preparation and Application of Environmental Friendly Materials(Jilin Normal University) Ministry of EducationChangchun 130103China

Given the urgency of organic pollutant removal and the low efficiency of advanced oxidation processes(AOPs),a novel Bi-ZFO/BMO-Vo photocatalyst was fabricated via the solvothermal method.A coupling system was constructed to combine photocatalysis with peroxymonosulfate(PMS)oxidation processes,which synergistically degrade organic ***-ZFO/BMO-Vo exhibited excellent photocatalytic per-formance,which could remove 100％RhB in 110 min and degrade 100％MG in 70 min,and 88％H-TC in 50 *** excellent catalytic performance of Bi-ZFO/BMO-Vo was not only attributed to the syner-gistic effect of PMS activation and photocatalysis,but also attributed to the SPR effect of Bi nanoparticles,electron capture of oxygen vacancies,and intense contact of Bi-ZFO/BMO-Vo *** active species capture experiments and EPR tests indicated that 1O2,SO4-·,OH,and O2-worked together for the RhB *** degradation intermediates of RhB were identified by *** on the exper-imental results,the band structure and Z-scheme charge transfer mechanism were *** evaluation indicated that Bi-ZFO/BMO-Vo-Vis/PMS could significantly reduce RhB *** efficient and stable catalyst is expected to be used in organic wastewater degradation and practical applications.

关键词： Z-scheme heterojunction Peroxymonosulfate Photocatalytic degradation Bi2MoO6

来源：评论

学校读者我要写书评

暂无评论

Research on Fault Detection of Traction System Based on Multi-task Learning

Research on Fault Detection of Traction System Based on Mult...

引用

International Conference on Artificial Intelligence and Autonomous Transportation, AIAT 2024

作者： Guo, Xingang Rao, Guibin Cheng, Chao College of Computer Science and Engineering Changchun University of Technology Changchun130012 China

ISBN: (纸本)9789819639601

Under the background of high-speed train production performance degradation, aiming at the problem of insufficient sensitivity of high-speed train traction system for early fault detection, we propose an early fault detection method based on multi-task learning. By using the multi-task network structure, this method obtains the probability density function of the data with the help of kernel density estimation and detects the fault from the data and the probability density function value. By comparing and analyzing the detection results of two fault detection indexes, Hotelling’s T2 statistic and Kullback-Leibler divergence statistic, the identifiability of fault indexes is enhanced. In addition, the adopted multi-task learning network structure enables the networks between two different tasks to share information effectively so that the performance of the overall network for fault detection is improved. Finally, the traction drive control system platform developed by Central South university is used to verify the effectiveness of the proposed method. © Beijing Paike Culture Commu. Co., Ltd. 2025.

关键词： Railroads

来源：评论

学校读者我要写书评

暂无评论

Structure network topology and broadband 2.7 μm luminescence in Er3+/Pr3+ co-doped TeO2-BaF2-YF3 glass

引用

Ceramics International 2025年第13期51卷 18086-18094页

作者： Shao, Zihao Bai, Yuerong Zhou, Dechun Leng, Zhuang Han, Kexuan Qin, Fengjie School of Materials Science and Engineering Changchun University of Science and Technology Jilin Changchun130022 China College of Materials Science and Engineering Jilin University China Chongqing Research Institute Changchun University of Science and Technology Chongqing401120 China

In this paper, Er3+/Pr3+ co-doped tellurite fluoride glass samples were prepared by high-temperature melting method, and the structural and optical properties of the glass samples were investigated and discussed by Raman spectroscopy, DSC mapping, absorption spectroscopy, and fluorescence spectroscopy with a view to obtaining a gain medium with good luminescence performance for the 2.7 μm band. The results of Raman spectroscopy show that the glass has a maximum phonon energy of 755 cm−1. The lower maximum phonon energy is attributed to the introduction of F−, which leads to the breakage of the Te-O-Te bond in the structure and the formation of two kinds of bonds, O-Te-F and Te-F-Te, which makes the structure of TeO4 gradually transformed into TeO3+1 and TeO3, and the fluoride–tellurite crystal system presents two phases of oxyfluoride (TeO3F2 and TeOF2) based on the formation of α and β TeO2 polymorphs. On the other hand, the change of structure network topology effectively inhibits the interaction between rare earth ions and reduces their nonradiative transitions probabilities, which provides a powerful environment for the mid-infrared band luminescence thus enhancing the infrared luminescence performance. The DSC pattern shows that the glass has a Tg of 355 °C and a ΔT as high as 118 °C, which suggests that the glass has a good thermal stability. The fluorescence spectra show that when the doping concentration of Er2O3/PrF3 is 1 mol% and 2 mol%, the maximum fluorescence intensity of 2.7 μm is acquired. The fluorescence half-height width of the glass is 138 nm, and the effective fluorescence bandwidth of the glass is Δλeff = 147.4 nm, which is a sufficient guarantee for the realization of the tunable laser of 2.7 μm with a wide Δλeff. The Judd-Ofelt intensity parameters were calculated by combining the absorption spectra, where the Ω2,Ω4,Ω6, and spectral quality factor (χ = Ω4/Ω6) of the glass were 6.83 × 10−20 cm2, 2.49 × 10−20 cm2,1.78 × 10−20 cm2, and 1.40, resp

关键词： Erbium compounds

来源：评论

学校读者我要写书评

暂无评论

Face Reconstruction-Based Generalized Deepfake Detection Model with Residual Outlook Attention

引用

ACM Transactions on Multimedia Computing, Communications and Applications 2025年第4期21卷 1-19页

作者： Shi, Zenan Liu, Wenyu Chen, Haipeng College of Computer Science and Technology Jilin University Changchun China Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education Jilin University Changchun China

With the continuous development of deep counterfeiting technology, the information security in our daily life is under serious threat. While existing face forgery detection methods exhibit impressive accuracy when applied to datasets such as FaceForensics++ and Celeb-DF, they falter significantly when confronted with out-of-domain scenarios. This causes specialization of learned representations to known forgery patterns presented in the training set, rendering it difficult to detect forgeries with unknown patterns. To address this challenge, we propose a novel end-to-end Face Reconstruction-Based Generalized Deepfake Detection (FRG2D) model with Residual Outlook Attention (ROA), which emphasizes the robust visual representations of genuine faces and discerns the subtle differences between authentic and manipulated facial images. Our methodology entails reconstructing authentic face images using an encoder-decoder architecture based on U-net, facilitating a deeper understanding of disparities between genuine and manipulated facial images. Furthermore, we integrate the convolutional block attention module (CBAM) and channel attention block (CAB) to selectively focus the network's attention on salient features within real face images. Furthermore, we employ ROA to guide the network's focus towards precise features within manipulated facial images. Simultaneously, the computed reconstruction differences obtained through ROA serves as the ultimate representation fed into the classifier for face forgery detection. Both the reconstruction and classification learning processes are optimized end-to-end. Through extensive experimentation, our model demonstrated a substantial improvement in deepfake detection across unknown domains, while maintaining a high accuracy within the known domain. © 2025 Copyright held by the owner/author(s). Publication rights licensed to ACM.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Variety classification and identification of maize seeds based on hyperspectral imaging method

引用

Optoelectronics Letters 2025年第4期21卷 234-241页

作者： XUE Hang XU Xiping MENG Xiang College of Optoelectronic Engineering Changchun University of Science and TechnologyChangchun 130022China College of Electronic and Information Engineering Beihua UniversityJilin 132021China

In this study,eight different varieties of maize seeds were used as the research *** 81 types of combined preprocessing on the original *** comparison,Savitzky-Golay(SG)-multivariate scattering correction(MSC)-maximum-minimum normalization(MN)was identified as the optimal preprocessing *** competitive adaptive reweighted sampling(CARS),successive projections algorithm(SPA),and their combined methods were employed to extract feature *** models based on back propagation(BP),support vector machine(SVM),random forest(RF),and partial least squares(PLS)were established using full-band data and feature *** all models,the(CARS-SPA)-BP model achieved the highest accuracy rate of 98.44%.This study offers novel insights and methodologies for the rapid and accurate identification of corn seeds as well as other crop seeds.

关键词： feature extraction extract feature wavelengthsclassification models variety classification hyperspectral imaging combined preprocessing competitive adaptive reweighted sampling cars successive projections algorithm spa preprocessing maize seeds

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：