检索结果-内蒙古大学图书馆

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

A data representation method using distance correlation

引用

Frontiers of Computer Science 2025年第1期19卷 1-14页

作者： Xinyan LIANG Yuhua QIAN Qian GUO keyin ZHENG Institute of Big Data Science and Industry Shanxi UniversityTaiyuan 030006China Key Laboratory of Computational Intelligence and Chinese Information Processing of Ministry of Education Shanxi UniversityTaiyuan 030006China School of Computer Science and Technology Taiyuan University of Science and TechnologyTaiyuan 030024China Shanxi Key Laboratory of Big Data Analysis and Parallel Computing Taiyuan University of Science and TechnologyTaiyuan 030024China

Association in-between features has been demonstrated to improve the representation ability of data. However, the original association data reconstruction method may face two issues: the dimension of reconstructed data is undoubtedly higher than that of original data, and adopted association measure method does not well balance effectiveness and efficiency. To address above two issues, this paper proposes a novel association-based representation improvement method, named as AssoRep. AssoRep first obtains the association between features via distance correlation method that has some advantages than Pearson’s correlation coefficient. Then an improved matrix is formed via stacking the association value of any two features. Next, an improved feature representation is obtained by aggregating the original feature with the enhancement matrix. Finally, the improved feature representation is mapped to a low-dimensional space via principal component analysis. The effectiveness of AssoRep is validated on 120 datasets and the fruits further prefect our previous work on the association data reconstruction.

关键词： association representation distance correlation classification

来源：评论

学校读者我要写书评

暂无评论

Improving Generalization in Offline Reinforcement Learning via Adversarial Data Splitting 41

Improving Generalization in Offline Reinforcement Learning v...

引用

41st International Conference on Machine Learning, ICML 2024

作者： Wang, Da Li, Lin Wei, Wei Yu, Qixian Hao, Jianye Liang, Jiye Key Laboratory of Computational Intelligence and Chinese Information Processing Ministry of Education School of Computer and Information Technology Shanxi University Taiyuan China College of Intelligence and Computing Tianjin University Tianjin China

Offline Reinforcement Learning (RL) commonly suffers from the out-of-distribution (OOD) overestimation issue due to the distribution shift. Prior work gradually shifts their focus from suppressing OOD overestimation to avoiding overly conservative learning from suboptimal behavior policies to improve generalization. However, most approaches explicitly delimit boundaries for OOD actions based on the support in the dataset, which can potentially impede the data near these boundaries from acquiring realistic estimates. This paper investigates how to loosen the rigid demarcation of OOD boundaries, adaptively extracting knowledge from empirical data to implicitly improve the model's generalization to nearby unseen data. We introduce an adversarial data splitting (ADS) framework that enforces the model to generalize the distribution shifts simulated from the train/validation subsets splitting of the dataset. Specifically, ADS is modeled as a min-max optimization problem inspired by meta-learning and solved by iterating over the following two steps. First, we train the model on the train-subset to minimize its loss on the validation-subset. Then, we adversarially generate the "hardest" train/validation subsets with the maximum distribution shift, making the model incapable of generalization at that splitting. We derive a generalization error bound for theoretically understanding ADS and verify the effectiveness with extensive experiments. Code is available at https://***/DkING-lv6/ADS. Copyright 2024 by the author(s)

关键词： Adversarial machine learning

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

An Empirical Study on Neural Networks Pruning: Trimming for Reducing Memory or Workload 5

An Empirical Study on Neural Networks Pruning: Trimming for ...

引用

5th International Conference on Data-Driven Optimization of Complex Systems, DOCS 2023

作者： Xiao, Kaiwen Cai, Xiaodong Xu, Ke School of Artificial Intelligence Anhui University Hefei China Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education Hefei China

ISBN: (纸本)9798350393521

Most of existing studies on neural network pruning only consider memory-based pruning strategies. However pruning for computational workload is often more important in hardware deployments due to a greater focus on model computation reductions. In addition, most pruning schemes restore model accuracy during pruning at the expense of adding hyperparameters, extending training time and training complexity. This work proposes a statistical-based globally soft iterative pruning scheme. With little extra calculation, an extremely sparse model can be obtained without additional hyperparameters and extended training time. Moreover, this work proposes the concept of computational intensity to balance model memory and computational workload during pruning. Focusing on memory orientated pruning, we can achieve 303×, 100× and 25× parameter compression on LeNet-5 (MNIST), VGG (CIFAR-10) and AlexNet (ImageNet) models, respectively. In particular, combined with cluster quantization, the LeNet-5 model parameters can be compressed to 3232×. Focusing on workload orientated pruning, we can reduce the computation by 7}.6× on the AlexNet model, without accuracy loss, significantly higher than prior work. In addition, in order to verify the versatility of the pruning method, we also migrate the pruning task to the object detection and implement 10× parameter compression and 2.8× computation compression for YOLOv2 with reduced mAP within 1%. © 2023 IEEE.

关键词： Iterative methods

来源：评论

学校读者我要写书评

暂无评论

An Update Method of Decision Implication Canonical Basis on Attribute Granulating

引用

Intelligent Automation & Soft Computing 2023年第8期37卷 1833-1851页

作者： Yanhui Zhai Rujie Chen Deyu Li School of Computer and Information Technology Shanxi UniversityTaiyuanShanxi Province030006China Key Laboratory of Computational Intelligence and Chinese Information Processing of Ministry of Education Shanxi UniversityTaiyuanShanxi Province030006China

Decision implication is a form of decision knowledge represen-tation,which is able to avoid generating attribute implications that occur between condition attributes and between decision *** with other forms of decision knowledge representation,decision implication has a stronger knowledge representation *** granularization may facilitate the knowledge extraction of different attribute granularity layers and thus is of application *** implication canonical basis(DICB)is the most compact set of decision implications,which can efficiently represent all knowledge in the decision *** order to mine all deci-sion information on decision context under attribute granulating,this paper proposes an updated method of *** this end,the paper reduces the update of DICB to the updates of decision premises after deleting an attribute and after adding granulation attributes of some *** on this,the paper analyzes the changes of decision premises,examines the properties of decision premises,designs an algorithm for incrementally generating DICB,and verifies its effectiveness through *** real life,by using the updated algorithm of DICB,users may obtain all decision knowledge on decision context after attribute granularization.

关键词： Decision context attribute granulating decision implication decision implication canonical basis

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

Denoising Diffusion Path: Attribution Noise Reduction with An Auxiliary Diffusion Model 38

Denoising Diffusion Path: Attribution Noise Reduction with A...

引用

38th Conference on Neural Information processing Systems, NeurIPS 2024

作者： Lei, Yiming Li, Zilong Zhang, Junping Shan, Hongming Shanghai Key Laboratory of Intelligent Information Processing School of Computer Science Fudan University China Institute of Science and Technology for Brain-Inspired Intelligence MOE Key Laboratory of Computational Neuroscience and Brain-Inspired Intelligence MOE Frontiers Center for Brain Science Fudan University China

The explainability of deep neural networks (DNNs) is critical for trust and reliability in AI systems. Path-based attribution methods, such as integrated gradients (IG), aim to explain predictions by accumulating gradients along a path from a baseline to the target image. However, noise accumulated during this process can significantly distort the explanation. While existing methods primarily concentrate on finding alternative paths to circumvent noise, they overlook a critical issue: intermediate-step images frequently diverge from the distribution of training data, further intensifying the impact of noise. This work presents a novel Denoising Diffusion Path (DDPath) to tackle this challenge by harnessing the power of diffusion models for denoising. By exploiting the inherent ability of diffusion models to progressively remove noise from an image, DDPath constructs a piece-wise linear path. Each segment of this path ensures that samples drawn from a Gaussian distribution are centered around the target image. This approach facilitates a gradual reduction of noise along the path. We further demonstrate that DDPath adheres to essential axiomatic properties for attribution methods and can be seamlessly integrated with existing methods such as IG. Extensive experimental results demonstrate that DDPath can significantly reduce noise in the attributions-resulting in clearer explanations-and achieves better quantitative results than traditional path-based methods. © 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

Evolutionary Multi-Objective Neural Architecture Search for Generalized Cognitive Diagnosis Models 5

Evolutionary Multi-Objective Neural Architecture Search for ...

引用

5th International Conference on Data-Driven Optimization of Complex Systems, DOCS 2023

作者： Yang, Shangshang Zhen, Cheng Tian, Ye Ma, Haiping Liu, Yuanchao Zhang, Panpan Zhang, Xingyi School of Artificial Intelligence Anhui University Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education Hefei China Institutes of Physical Science and Information Technology Anhui University Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education Hefei China Automation for Process Industries Northeastern University State Key Laboratory of Synthetical Shenyang China

ISBN: (纸本)9798350393521

Cognitive diagnosis models (CDMs) with high generalization are essential for intelligent education systems to reveal students' knowledge states in multiple datasets. However, existing CDMs' architectures are designed dependent on researcher expertise and experience from observing and summarizing partial students' learning behaviors, which makes handcrafted models not cover all learning behaviors well and thus limits their generalization. To develop generalized CDMs, this paper proposes an evolutionary neural architecture search to design CDMs' architectures effective on multiple datasets automatically. Specifically, we first formulate the search task as a multi-objective optimization problem (MOP), which maximizes model performance on multiple datasets containing learning behaviors as many as possible to ensure model generalization. Then, an expressive search space is devised to contain as many potential architectures as possible, where each architecture is denoted by a unified form, taking three given inputs and integrating them in a linear or no-linear manner for prediction. Finally, an evolutionary algorithm with a tailored deduplication strategy is employed to solve the MOP, where each architecture is further represented by a single-root computation tree for easy optimization. Experiments on multiple datasets demonstrate the generalization and effectiveness of the best architecture searched by the proposed approach, and the searched architecture also holds as good interpretability as handcrafted architectures. © 2023 IEEE.

关键词： Multiobjective optimization

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：