检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Jayatilleke, Nevidu Weerasinghe, Ruvan School of Computing Informatics Institute of Technology Colombo Sri Lanka

Automatic patent summarization approaches that help in the patent analysis and comprehension procedure are in high demand due to the colossal growth of innovations. The development of natural language processing (NLP), text mining, and deep learning has notably amplified the efficacy of text summarization models for abundant types of documents. Summarizing patent text remains a pertinent challenge due to the labyrinthine writing style of these documents, which includes technical and legal intricacies. Additionally, these patent document contents are considerably lengthier than archetypal documents, which intricates the process of extracting pertinent information for summarization. Embodying extractive and abstractive text summarization methodologies into a hybrid framework, this study proposes a system for efficiently creating abstractive summaries of patent records. The procedure involves leveraging the LexRank graph-based algorithm to retrieve the important sentences from input parent texts, then utilizing a Bidirectional Auto-Regressive Transformer (BART) model that has been fine-tuned using Low-Ranking Adaptation (LoRA) for producing text summaries. This is accompanied by methodical testing and evaluation strategies. Furthermore, the author employed certain meta-learning techniques to achieve Domain Generalization (DG) of the abstractive component across multiple patent fields. Copyright © 2025, The Authors. All rights reserved.

关键词： Patents and inventions

来源：评论

学校读者我要写书评

暂无评论

Incentive-Compatible Federated Learning with Stackelberg Game Modeling

arXiv

引用

arXiv 2025年

作者： Javaherian, Simin Turney, Bryce Chen, Li Tzeng, Nian-Feng School of Computing and Informatics University of Louisiana Lafayette United States

Federated Learning (FL) has gained prominence as a decentralized machine learning paradigm, allowing clients to collaboratively train a global model while preserving data privacy. Despite its potential, FL faces significant challenges in heterogeneous environments, where varying client resources and capabilities can undermine overall system performance. Existing approaches primarily focus on maximizing global model accuracy, often at the expense of unfairness among clients and suboptimal system efficiency, particularly in non-IID (nonIndependent and Identically Distributed) settings. In this paper, we introduce FLamma, a novel Federated Learning framework based on adaptive gamma-based Stackelberg game, designed to address the aforementioned limitations and promote fairness. Our approach allows the server to act as the leader, dynamically adjusting a decay factor while clients, acting as followers, optimally select their number of local epochs to maximize their utility. Over time, the server incrementally balances client influence, initially rewarding higher-contributing clients and gradually leveling their impact, driving the system toward a Stackelberg Equilibrium. Extensive simulations on both IID and non-IID datasets show that our method significantly improves fairness in accuracy distribution without compromising overall model performance or convergence speed, outperforming traditional FL baselines. Copyright © 2025, The Authors. All rights reserved.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

SSL400 - A Comprehensive Word Level Dataset for Sinhala Sign Language Recognition

SSL400 - A Comprehensive Word Level Dataset for Sinhala Sign...

引用

Advanced Research in computing (ICARC), International Conference on

作者： Yohan Abhishek T. G. D. K. Sumanathilaka School of Computing Informatics Institute of Technology Colombo Sri Lanka

ISBN: (数字)9798331530983

ISBN: (纸本)9798331530990

In the context of increasing digital accessibility and the need for inclusive communication technologies, the demand for automated sign language recognition has grown significantly. However, Sinhala Sign Language (SSL), used by the Sri Lankan deaf community, remains underexplored due to the lack of labelled datasets and the challenges of developing recognition models for low-resource languages. This study addresses this gap by introducing an SSL dataset, specifically curated to support real-time recognition of 384 commonly used words in daily and work-related contexts. The methodology for dataset creation involved expert consultations, a structured data collection process, and multi-step labelling. These steps ensured the accurate labeling of video data, which can be used for machine learning tasks like gesture recognition, sentence formation, and speech translation. The proposed dataset lays the foundation for SSL recognition models that are fast, lightweight, and optimized for low-resource environments. Additionally, we explore the performance of initial recognition models using this dataset, demonstrating its applicability in real-world scenarios. The dataset and research outcomes provide valuable resources for the development of inclusive AI-driven communication tools for the Sinhala-speaking deaf community. This contribution addresses the pressing need for better technological support for underrepresented languages in the field of sign language recognition and processing.

关键词： Sign language Translation Speech recognition Pressing Machine learning Streaming media Data collection Real-time systems Communications technology Labeling

来源：评论

学校读者我要写书评

暂无评论

Systematic Review of Fine-tuning Approaches for Large Language Models in the Context of Sinhala

Systematic Review of Fine-tuning Approaches for Large Langua...

引用

Advanced Research in computing (ICARC), International Conference on

作者： Sachin Hansaka Deshan Sumanathilaka School of Computing Informatics Institute of Technology Colombo Sri Lanka

ISBN: (数字)9798331530983

ISBN: (纸本)9798331530990

The rapid development of large language models (LLMs) such as GPT-3, GPT-4, LlaMA, and mBERT has significantly advanced the natural language processing (NLP) field across many widely spoken languages. However, the effectiveness of these LLMs on low-resource languages such as Sinhala remained unexplored. With the scarcity of high-quality data and linguistic resources, fine-tuning LLMs in Sinhala has gained significant challenges. With a focus on Sinhala specifically, this systematic review addresses the methodologies and outcomes of adapting large language models for underrepresented languages. The purpose of this review is to synthesize existing research, identify effective strategies, contributions, and improvements, and highlight the challenges and opportunities in this domain. By analyzing various approaches, including data augmentation, transfer learning, cross-lingual transfer, parameter-efficient fine-tuning, adapter tuning, and model adaptation likewise, this review provides a comprehensive and precise idea of the current state of LLM fine-tuning on low-resource languages like Sinhala. The results highlight the importance of using creative methods and collaboration to improve LLM performance and usefulness in linguistic environments with limited resources.

关键词： Translation Large language models Transfer learning Collaboration Linguistics Data augmentation Natural language processing Tuning Systematic literature review Synthetic data

来源：评论

学校读者我要写书评

暂无评论

AGCL: Aspect Graph Construction and Learning for Aspect-level Sentiment Classification 31

AGCL: Aspect Graph Construction and Learning for Aspect-leve...

引用

31st International Conference on Computational Linguistics, COLING 2025

作者： Jian, Zhongquan Wu, Daihang Wang, Shaopan Wang, Yancheng Yao, Junfeng Wang, Meihong Wu, Qingqiang Institute of Artificial Intelligence Xiamen University China School of Informatics Xiamen University China School of Film Xiamen University China College of Management Mahidol University Thailand School of Computing and Data Science Xiamen University Malaysia Malaysia Key Laboratory of Digital Protection and Intelligent Processing of Intangible Cultural Heritage of Fujian and Taiwan Ministry of Culture and Tourism Xiamen University China Xiamen Key Laboratory of Intelligent Storage and Computing School of Informatics Xiamen University China

ISBN: (纸本)9798891761964

Prior studies on Aspect-level Sentiment Classification (ALSC) emphasize modeling interrelationships among aspects and contexts but overlook the crucial role of aspects themselves as essential domain knowledge. To this end, we propose AGCL, a novel Aspect Graph Construction and Learning method, aimed at furnishing the model with finely tuned aspect information to bolster its task-understanding ability. AGCL's pivotal innovations reside in Aspect Graph Construction (AGC) and Aspect Graph Learning (AGL), where AGC harnesses intrinsic aspect connections to construct the domain aspect graph, and then AGL iteratively updates the introduced aspect graph to enhance its domain expertise, making it more suitable for the ALSC task. Hence, this domain aspect graph can serve as a bridge connecting unseen aspects with seen aspects, thereby enhancing the model's generalization capability. Experiment results on three widely used datasets demonstrate the significance of aspect information for ALSC and highlight AGL's superiority in aspect learning, surpassing state-of-the-art baselines greatly. Code is available at https://***/jian-projects/agcl. © 2025 Association for Computational Linguistics.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Advancements in Natural Language Processing for Automatic Text Summarization

arXiv

引用

arXiv 2025年

作者： Jayatilleke, Nevidu Weerasinghe, Ruvan Senanayake, Nipuna School of Computing Informatics Institute of Technology Colombo Sri Lanka

The substantial growth of textual content in diverse domains and platforms has led to a considerable need for Automatic Text Summarization (ATS) techniques that aid in the process of text analysis. The effectiveness of text summarization models has been significantly enhanced in a variety of technical domains because of advancements in Natural Language Processing (NLP) and Deep Learning (DL). Despite this, the process of summarizing textual information continues to be significantly constrained by the intricate writing styles of a variety of texts, which involve a range of technical complexities. Text summarization techniques can be broadly categorized into two main types: abstractive summarization and extractive summarization. Extractive summarization involves directly extracting sentences, phrases, or segments of text from the content without making any changes. On the other hand, abstractive summarization is achieved by reconstructing the sentences, phrases, or segments from the original text using linguistic analysis. Through this study, a linguistically diverse categorizations of text summarization approaches have been addressed in a constructive manner. In this paper, the author explored existing hybrid techniques that have employed both extractive and abstractive methodologies. In addition, the pros and cons of various approaches discussed in the literature are also investigated. Furthermore, a comparative analysis conducted to evaluate the generated summaries utilizing language generation models, employing different techniques and matrices. This survey endeavors to provide a comprehensive overview of ATS by presenting the progression of language processing regarding this task through a breakdown of diverse systems and architectures accompanied by technical and mathematical explanations of their operations. Finally, following the conclusion, a discussion is held regarding the limitations, challenges, and potential future work that can be conducted within this

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

An Examination of Harmonizing Multi-Vendor Cardiac MRI Data for Vendor-Neutral Image Segmentation

An Examination of Harmonizing Multi-Vendor Cardiac MRI Data ...

引用

Advanced Research in computing (ICARC), International Conference on

作者： Hamdaan Mohideen Mohamed Ayoob Oshan Nettasinghe Vithushan Sylvester Helmini Bowala School of Computing Informatics Institute of Technology Colombo Sri Lanka

ISBN: (数字)9798331530983

ISBN: (纸本)9798331530990

Vendor-specific scanner hardware, software, and image acquisition protocols can introduce systematic biases into MRI scans. This introduces non-biological artifacts which introduce variations across MRI images from different vendors. Further, we ratified such non-biological variances by conducting a discrepancy study on the MnMs-2 Cardiac MRI dataset. Data harmonization is the process of combining datasets from different sources while removing unwanted variations that are not caused by biological differences. This is important because it allows researchers to increase the understanding of biological differences in data, while mitigating vendor-specific biases. We used a domain adaptation methodology (which is a data harmonization method) on the dataset to unlearn domain (or vendor) specific biases by introducing a confusion loss to confuse the encoder. Furthermore, we introduced a 3D segmentation model to this method to capture spatial information. This method demonstrates the performance on the main task (segmentation) and compares the domain predictor, before and after unlearning. This shows the magnitude of mitigating vendor-specific biases of the model. In addition to the model architecture performance, we documented the discrepancy study between inter-vendor model performance to demonstrate the magnitude of impact in which vendor-specific, and non-biological variations of MRI scans can affect the model.

关键词： Training Image segmentation Analytical models Adaptation models Three-dimensional displays Accuracy Magnetic resonance imaging Computational modeling Data models Biological information theory

来源：评论

学校读者我要写书评

暂无评论

CRP Encapsulation Scheme to Protect Medical Record Stored in the Cloud

CRP Encapsulation Scheme to Protect Medical Record Stored in...

引用

IEEE International Conference on Consumer Electronics (ICCE)

作者： Dina Ghanai Miandoab Adil Alshammari Bertrand Cambou Sareh Assiri School of Informatics and Computing Northern Arizona University Flagstaff Arizona

ISBN: (数字)9798331521165

ISBN: (纸本)9798331521172

This paper presents protocols allowing patients to control access to their medical files without complex key management. Several priorities were considered in this work: ease of use, fast latencies, uncompromised privacy protection, confidentiality, and an initial focus on Electronic Health Records (EHR) and Electronic Medical Records (EMR). We implemented multi-faceted protections in which: multiple parties can have control over each file, each file is encrypted with its own key, and partially homomorphic verification methods are applied to the encrypted files. The proposed protocols strike a balance between making medical data accessible and preserving privacy, thus enhancing data security and trust in digital healthcare systems.

关键词： Encapsulation Data privacy Privacy Protocols Authentication Electronic healthcare Cryptography Electronic medical records Protection Consumer electronics

来源：评论

学校读者我要写书评

暂无评论

FarExStance: Explainable Stance Detection for Farsi 31

FarExStance: Explainable Stance Detection for Farsi

引用

31st International Conference on Computational Linguistics, COLING 2025

作者： Zarharan, Majid Hashemi, Maryam Behroozrazegh, Malika Eetemadi, Sauleh Pilehvar, Mohammad Taher Foster, Jennifer School of Computing Dublin City University Ireland Iran University of Science and Technology Iran Dadmatech Tehran Iran School of Computer Science University of Birmingham United Kingdom School of Computer Science and Informatics Cardiff University United Kingdom

ISBN: (纸本)9798891761964

We introduce FarExStance, a new dataset for explainable stance detection in Farsi. Each instance in this dataset contains a claim, the stance of an article or social media post towards that claim, and an extractive explanation which provides evidence for the stance label. We compare the performance of a fine-tuned multilingual RoBERTa model to several large language models in zero-shot, few-shot, and parameter-efficient fine-tuned settings on our new dataset. On stance detection, the most accurate models are the fine-tuned RoBERTa model, the LLM Aya-23-8B which has been fine-tuned using parameter-efficient fine-tuning, and few-shot Claude-3.5-Sonnet. Regarding the quality of the explanations, our automatic evaluation metrics indicate that few-shot GPT-4o generates the most coherent explanations, while our human evaluation reveals that the best Overall Explanation Score (OES) belongs to few-shot Claude-3.5-Sonnet. The fine-tuned Aya-32-8B model produced explanations most closely aligned with the reference explanations. © 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

Simple One-step Multi-view Clustering with Fast Similarity and Cluster Structure Learning

引用

IEEE Signal Processing Letters 2025年 32卷 1850-1854页

作者： Xia, Xianxian Huang, Dong Yang, Chen-Min He, Chaobo Wang, Chang-Dong South China Agricultural University College of Mathematics and Informatics Guangzhou China South China Normal University School of Computer Science Guangzhou China Sun Yat-sen University School of Computer Science and Engineering Guangzhou China Ministry of Education Key Laboratory of Machine Intelligence and Advanced Computing China

Multi-view clustering (MVC) is essential for integrating heterogeneous data from multiple sources. However, many existing approaches are hindered by high computational complexity and the separate optimization of similarity and cluster structures. In light of these challenges, this paper presents a novel anchor-based MVC method termed simple one-step multi-view clustering with fast similarity and cluster structure learning (SONIC), which models adaptive anchor learning, multi-view similarity structure learning, and discrete cluster structure learning in a joint framework. In particular, we employ the anchor-based multi-view similarity learning to capture the consensus manifold structure latent in multiple views, thereby constructing a unified bipartite graph with adaptive anchor learning and view weighting. Then we impose a low-rank constraint on the bipartite graph structure to directly reveal the desired number of clusters without additional post-processing. An efficient alternating minimization algorithm is developed to optimize the model, resulting in a computational complexity that scales linearly with the number of samples. Extensive experiments on eight benchmark datasets demonstrate the superior performance of SONIC in both clustering quality and computational efficiency. © 1994-2012 IEEE.

关键词： Data assimilation

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：