检索结果-内蒙古大学图书馆

41st International Conference on Machine Learning, ICML 2024

作者： Liang, Jianqing Chen, Min Liang, Jiye Key Laboratory of Computational Intelligence and Chinese Information Processing of Ministry of Education School of Computer and Information Technology Shanxi University Shanxi Taiyuan030006 China

The Transformer architecture has recently gained considerable attention in the field of graph representation learning, as it naturally overcomes several limitations of Graph Neural Networks (GNNs) with customized attention mechanisms or positional and structural encodings. Despite making some progress, existing works tend to overlook external information of graphs, specifically the correlation between graphs. Intuitively, graphs with similar structures should have similar representations. Therefore, we propose Graph External Attention (GEA) - a novel attention mechanism that leverages multiple external node/edge key-value units to capture inter-graph correlations implicitly. On this basis, we design an effective architecture called Graph External Attention Enhanced Transformer (GEAET), which integrates local structure and global interaction information for more comprehensive graph representations. Extensive experiments on benchmark datasets demonstrate that GEAET achieves state-of-the-art empirical performance. The source code is available for reproducibility at: https://***/icm1018/GEAET. Copyright 2024 by the author(s)

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

A Bayesian matrix factorization model for dynamic user embedding in recommender system

引用

Frontiers of Computer Science 2022年第5期16卷 233-235页

作者： Kaihan ZHANG Zhiqiang WANG Jiye LIANG Xingwang ZHAO Key Laboratory of Computational Intelligence and Chinese Information Processing of Ministry of Education School of Computer and Information TechnologyShanxi UniversityTaiyuan 030006China

1 Introduction The main idea of recommender system is how to learn accurate users’embeddings from behavior data[1].Each dimension of users’embeddings can reflect the interests of users in different potential *** real-world scenarios,users’interests are drifting over time,which brings a challenge to learn accurate dynamic users’***,various time-aware recommendation methods have been proposed to tackle this problem by modeling the evolution process of users’interests[2−4].However,they usually assume that users’embeddings drift with the same range on all *** practice,users’embeddings should change diversely on different dimensions over ***,for the rapidly changing interests of the users,the corresponding dimensions should change *** the contrary,the dimensions representing stable interests may change slightly.

关键词： dimensions system dynamic

来源：评论

学校读者我要写书评

暂无评论

LOG: A Local-to-Global Optimization Approach for Retrieval-based Explainable Multi-Hop Question Answering 31

LOG: A Local-to-Global Optimization Approach for Retrieval-b...

引用

31st International Conference on Computational Linguistics, COLING 2025

作者： Xu, Hao Zhao, Yunxiao Zhang, Jiayang Wang, Zhiqiang Li, Ru School of Computer and Information Technology Shanxi University Taiyuan China Key Laboratory of Computational Intelligence and Chinese Information Processing of Ministry of Education Shanxi University Taiyuan China

ISBN: (纸本)9798891761964

Multi-hop question answering (MHQA) aims to utilize multi-source intensive documents retrieved to derive the answer. However, it is very challenging to model the importance of knowledge retrieved. Previous approaches primarily emphasize single-step and multi-step iterative decomposition or retrieval, which are susceptible to failure in long-chain reasoning due to the progressive accumulation of erroneous information. To address this problem, we propose a novel Local-tO-Global optimized retrieval method (LOG) to discover more beneficial information, facilitating the MHQA. In particular, we design a pointwise conditional V-information based local information modeling to cover usable documents with reasoning knowledge. We also improve tuplet objective loss, advancing multi-examples-aware global optimization to model the relationship between scattered documents. Extensive experimental results demonstrate our proposed method outperforms prior state-of-the-art models, and it can significantly improve multi-hop reasoning, notably for long-chain reasoning. Our code is available at https://***/yunxiaomr/LOG. © 2025 Association for Computational Linguistics.

关键词： Question answering

来源：评论

学校读者我要写书评

暂无评论

Enhancing Event Causality Identification with LLM Knowledge and Concept-Level Event Relations 31

Enhancing Event Causality Identification with LLM Knowledge ...

引用

31st International Conference on Computational Linguistics, COLING 2025

作者： Su, Ya Zhang, Hu Zhang, Guangjun Wang, Yujie Fan, Yue Li, Ru Wang, Yuanlong School of Computer and Information Technology Shanxi University Taiyuan China Key Laboratory of Computational Intelligence and Chinese Information Processing of Ministry of Education Shanxi University Taiyuan China

ISBN: (纸本)9798891761964

Event Causality Identification (ECI) aims to identify fine-grained causal relationships between events in an unstructured text. Existing ECI methods primarily rely on knowledge-enhanced and graph-based reasoning approaches, but they often overlook the dependencies between similar events. Additionally, the connection between unstructured text and structured knowledge is relatively weak. Therefore, this paper proposes an ECI method enhanced by LLM Knowledge and Concept-Level Event Relations (LKCER). Specifically, LKCER constructs a conceptual-level heterogeneous event graph by leveraging the local contextual information of related event mentions, generating a more comprehensive global semantic representation of event concepts. At the same time, the knowledge generated by COMET is filtered and enriched using LLM, strengthening the associations between event pairs and knowledge. Finally, the joint event conceptual representation and knowledge-enhanced event representation are used to uncover potential causal relationships between events. The experimental results show that our method outperforms previous state-of-the-art methods on both benchmarks, EventStoryLine and Causal-TimeBank. © 2025 Association for Computational Linguistics.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Lightweight Food Image Recognition With Global Shuffle Convolution

IEEE Transactions on AgriFood Electronics

引用

IEEE Transactions on AgriFood Electronics 2024年第2期2卷 392-402页

作者： Sheng, Guorui Min, Weiqing Yao, Tao Song, Jingru Yang, Yancun Wang, Lili Jiang, Shuqiang Ludong University Department of Information and Electrical Engineering Yantai264025 China Institute of Computing Technology Chinese Academy of Sciences Key Laboratory of Intelligent Information Processing Beijing100190 China University of Chinese Academy of Sciences Beijing100049 China

Consumer behaviors and habits in food choices impact their physical health and have implications for climate change and global warming. Efficient food image recognition can assist individuals in making more environmentally friendly and healthier dietary choices using end devices, such as smartphones. Simultaneously, it can enhance the efficiency of server-side training, thereby reducing carbon emissions. We propose a lightweight deep neural network named Global Shuffle Net (GSNet) that can efficiently recognize food images. In GSNet, we develop a novel convolution method called global shuffle convolution, which captures the dependence between long-range pixels. Merging global shuffle convolution with classic local convolution yields a framework that works as the backbone of GSNet. Through GSNet's ability to capture the dependence between long-range pixels at the start of the network, by restricting the number of layers in the middle and rear, the parameters and floating operation operations (FLOPs) can be minimized without compromising the performance, thus permitting a lightweight goal to be achieved. Experimental results on four popular food recognition datasets demonstrate that our approach achieves state-of-the-art performance with higher accuracy and fewer FLOPs and parameters. For example, in comparison to the current state-of-the-art model of MobileViTv2, GSNet achieved 87.9% accuracy of the top-1 level on the Eidgenössische Technische Hochschule Zürich (ETHZ) Food-101 dataset with 28% reduction in the parameters, 37% reduction in the FLOPs, but a 0.7% more accuracy. © 2024 IEEE.

关键词： Carbon dioxide

来源：评论

学校读者我要写书评

暂无评论

Iteratively Refined Behavior Regularization for Offline Reinforcement Learning 38

Iteratively Refined Behavior Regularization for Offline Rein...

引用

38th Conference on Neural information processing Systems, NeurIPS 2024

作者： Ma, Yi Hao, Jianye Hu, Xiaohan Zheng, Yan Xiao, Chenjun School of Computer and Information Technology Shanxi University China Key Laboratory of Computational Intelligence and Chinese Information Processing of Ministry of Education China College of Intelligence and Computing Tianjin University China Noah's Ark Lab Huawei Canada The Chinese University of Hong Kong Shenzhen China

One of the fundamental challenges for offline reinforcement learning (RL) is ensuring robustness to data distribution. Whether the data originates from a near-optimal policy or not, we anticipate that an algorithm should demonstrate its ability to learn an effective control policy that seamlessly aligns with the inherent distribution of offline data. Unfortunately, behavior regularization, a simple yet effective offline RL algorithm, tends to struggle in this regard. In this paper, we propose a new algorithm that substantially enhances behavior-regularization based on conservative policy iteration. Our key observation is that by iteratively refining the reference policy used for behavior regularization, conservative policy update guarantees gradually improvement, while also implicitly avoiding querying out-of-sample actions to prevent catastrophic learning failures. We prove that in the tabular setting this algorithm is capable of learning the optimal policy covered by the offline dataset, commonly referred to as the in-sample optimal policy. We then explore several implementation details of the algorithm when function approximations are applied. The resulting algorithm is easy to implement, requiring only a few lines of code modification to existing methods. Experimental results on the D4RL benchmark indicate that our method outperforms previous state-of-the-art baselines in most tasks, clearly demonstrate its superiority over behavior regularization. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Research on mixed decision implications based on formal concept analysis

引用

International Journal of Cognitive computing in Engineering 2023年第1期4卷 71-77页

作者： Ren, Xingguo Li, Deyu Zhai, Yanhui School of Computer and Information Technology Shanxi University Shanxi Taiyuan030006 China Key Laboratory of Computational Intelligence and Chinese Information Processing of Ministry of Education Shanxi University Shanxi Taiyuan030006 China

Decision implication is an important form of knowledge representation and acquisition in Formal Concept Analysis. Decision implication reduces the redundancy of knowledge extracted from data. However, decision implication cannot extract negative information from data, so there is information loss in decision implication. This paper introduces negative attributes to decision implication and proposes mixed decision implication, enabling decision implication to extract negative knowledge from data and to represent richer decision knowledge. This paper studies the logical systems of mixed decision implications. The semantical system of mixed decision implications is constructed to represent and deduce sound mixed decision implications and avoid contradictory mixed decision implications. In the syntactical system, Mixed Augmentation and Mixed Combination are introduced and the soundness, completeness and non-redundancy of these two inference rules are proved. © 2023

关键词： Formal concept analysis

来源：评论

学校读者我要写书评

暂无评论

Extending generalized unsupervised manifold alignment

引用

Science China(information Sciences) 2022年第7期65卷 139-156页

作者： Xiaoyi YIN Zhen CUI Hong CHANG Bingpeng MA Shiguang SHAN Key Laboratory of Intelligent Information Processing of Chinese Academy of Sciences (CAS) Institute of Computing Technology Chinese Academy of Sciences University of Chinese Academy of Sciences School of Computer Science and Engineering Nanjing University of Science and Technology CAS Center for Excellence in Brain Science and Intelligence Technology

Building connections between different data sets is a fundamental task in machine learning and related application community. With proper manifold alignment, the correspondences between data sets will assist us with comprehensive study of data processes and analyses. Despite the several progresses in semi-supervised and unsupervised scenarios, potent manifold alignment methods in generalized and realistic circumstances remain in absence. Besides, theretofore unsupervised algorithms seldom prove themselves mathematically. In this paper, we devise an efficient method to properly solve the unsupervised manifold alignment problem and denominate it as extending generalized unsupervised manifold alignment(EGUMA)method. More specifically, an explicit relaxed integer programming method is adopted to solve the unsupervised manifold alignment problem, which reconciles three factors covering the updated local structure matching, the the feature comparability and geometric preservation. An additional effort is retained on extending the Frank Wolfe algorithm to tacking our optimization problem. Besides our previous endeavors we adopt a new strategy for neighborhood discovery in the manifolds. The main advantages over previous methods accommodate(1) simultaneous alignment and discovery of manifolds;(2) complete unsupervised learning structure without any prerequisite correspondence;(3) more concise local geometry for the embedding space;(4) efficient alternative optimization;(5) strict mathematical analysis on the convergence and efficiency issues. Experiments on real-world applications verify the high accuracy and efficiency of our proposed method.

关键词： unsupervised manifold alignment relaxed integer programming Frank Wolfe algorithm

来源：评论

学校读者我要写书评

暂无评论

Mitigating Shortcut Learning via Smart Data Augmentation based on Large Language Model 31

Mitigating Shortcut Learning via Smart Data Augmentation bas...

引用

31st International Conference on Computational Linguistics, COLING 2025

作者： Sun, Xinyi Tan, Hongye Guo, Yaxin Qiang, Pengpeng Li, Ru Zhang, Hu School of Computer and Information Technology Shanxi University Taiyuan China Institute of Intelligent Information Processing Shanxi University Taiyuan China Key Laboratory of Computational Intelligence and Chinese Information Processing of Ministry of Education Shanxi University Taiyuan China

ISBN: (纸本)9798891761964

Data-driven pre-trained language models typically perform shortcut learning wherein they rely on the spurious correlations between the data and the ground truth. This reliance can undermine the robustness and generalization of the model. To address this issue, data augmentation emerges as a promising solution. By integrating anti-shortcut data to the training set, the models' shortcut-induced biases can be mitigated. However, existing methods encounter three challenges: (1) Manual definition of shortcuts is tailored to particular datasets, restricting generalization. (2) The inherent confirmation bias during model training hampers the effectiveness of data augmentation. (3) Insufficient exploration of the relationship between the model performance and the augmented data quantity may result in excessive data consumption. To tackle these challenges, we propose a method of Smart Data Augmentation based on Large Language Models (SAug-LLM). It leverages the LLMs to autonomously identify shortcuts and generate their anti-shortcut counterparts. In addition, the dual validation is employed to mitigate the confirmation bias during the model retraining. Furthermore, the data augmentation process is optimized to effectively rectify model biases while minimizing data consumption. We validate the effectiveness and generalization of our method through extensive experiments across various natural language processing tasks, demonstrating an average performance improvement of 5.61%. © 2025 Association for Computational Linguistics.

关键词： Data assimilation

来源：评论

学校读者我要写书评

暂无评论

Speculative Decoding with CTC-based Draft Model for LLM Inference Acceleration 38

Speculative Decoding with CTC-based Draft Model for LLM Infe...

引用

38th Conference on Neural information processing Systems, NeurIPS 2024

作者： Wen, Zhuofan Gui, Shangtong Feng, Yang Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences China State Key Lab of Processors Institute of Computing Technology Chinese Academy of Sciences China Key Laboratory of Al Safety Chinese Academy of Sciences China University of Chinese Academy of Sciences Beijing China

Inference acceleration of large language models (LLMs) has been put forward in many application scenarios and speculative decoding has shown its advantage in addressing inference acceleration. Speculative decoding usually introduces a draft model to assist the base LLM where the draft model produces drafts and the base LLM verifies the draft for acceptance or rejection. In this framework, the final inference speed is decided by the decoding speed of the draft model and the acceptance rate of the draft provided by the draft model. Currently the widely used draft models usually generate draft tokens for the next several positions in a non-autoregressive way without considering the correlations between draft tokens. Therefore, it has a high decoding speed but an unsatisfactory acceptance rate. In this paper, we focus on how to improve the performance of the draft model and aim to accelerate inference via a high acceptance rate. To this end, we propose a CTC-based draft model which strengthens the correlations between draft tokens during the draft phase, thereby generating higher-quality draft candidate sequences. Experiment results show that compared to strong baselines, the proposed method can achieve a higher acceptance rate and hence a faster inference speed. © 2024 Neural information processing systems foundation. All rights reserved.

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：