检索结果-内蒙古大学图书馆

1st Workshop on Natural Language Processing for Indo-Aryan and Dravidian Languages, IndoNLP 2025

作者： Sritharan, Braveenan Thayasivam, Uthayasanker Dept. of Computer Science & Engineering University of Moratuwa Sri Lanka

ISBN: (纸本)9798891762145

Multilingual speaker identification and verification is a challenging task, especially for languages with diverse acoustic and linguistic features such as Indo-Aryan and Dravidian languages. Previous models have struggled to generalize across multilingual environments, leading to significant performance degradation when applied to multiple languages. In this paper, we propose an advanced approach to multilingual speaker identification and verification, specifically designed for Indo-Aryan and Dravidian languages. Empirical results on the Kathbath dataset show that our approach significantly improves speaker identification accuracy, reducing the performance gap between monolingual and multilingual systems from 15% to just 1%. Additionally, our model reduces the equal error rate for speaker verification from 15% to 5% in noisy conditions. Our method demonstrates strong generalization capabilities across diverse languages, offering a scalable solution for multilingual voice-based biometric systems. © 2025 Association for Computational Linguistics.

关键词： Speech recognition

来源：评论

学校读者我要写书评

暂无评论

Empirical Study of Zero-shot Keyphrase Extraction with Large Language Models 31

Empirical Study of Zero-shot Keyphrase Extraction with Large...

引用

31st International Conference on Computational Linguistics, COLING 2025

作者： Kang, Byungha Shin, Youhyun Department of Computer Science and Engineering Incheon National University Korea Republic of

ISBN: (纸本)9798891761964

This study investigates the effectiveness of Large Language Models (LLMs) for zero-shot keyphrase extraction (KE). We propose and evaluate four prompting strategies: vanilla, role prompting, candidate-based prompting, and hybrid prompting. Experiments conducted on six widely-used KE benchmark datasets demonstrate that Llama3-8B-Instruct with vanilla prompting outperforms state-of-the-art unsupervised methods, PromptRank, by an average of 9.43%, 7.68%, and 4.82% in F1@5, F1@10, and F1@15, respectively. Hybrid prompting, which combines the strengths of vanilla and candidate-based prompting, further enhances overall performance. Moreover role prompting, which assigns a task-related role to LLMs, consistently improves performance across various prompting strategies. We also explore the impact of model size and different LLM series: GPT-4o, Gemma2, and Qwen2. Results show that Llama3 and Gemma2 demonstrate the strongest zero-shot KE performance, with hybrid prompting consistently enhancing results across most LLMs. We hope this study provides insights to researchers exploring LLMs in KE tasks, as well as practical guidance for model selection in real-world applications. Our code is available at https://***/kangnlp/Zero-shot-KPE-with-LLMs. © 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

EventSync: Event Scheduler & Coordination Planner

EventSync: Event Scheduler & Coordination Planner

引用

2025 IEEE International Students' Conference on Electrical, Electronics and computer science, SCEECS 2025

作者： Saraswat, Mohit Kumar, Nishant Mishra, Pranav Kaur, Mandeep Sharda University Department of Computer Science & Engineering Greater Noida India

ISBN: (纸本)9798331529833

EventSync will change the doings in university events organizing. It brings in one-stop smooth workflow, thus bringing in flexibility and efficiency. EventSync presents an intuitive interface of bringing together all three scheduling, task management, and budget tracking in one place. All this reduces tensions and headaches in hand-coordinating and leaves more room for creativity and working together instead of doing work. It further quickens the approval processes;thus, the events can get their way from conceptualization to execution very quickly. Post-Event reporting by EventSync's helps in refining future events with data and feedback from stakeholders. The best optimization of resources, including coordination among team members, ensures that every event is carried out perfectly to create long-lasting impressions among customers. Therefore, when this innovative tool is taken aboard any University, event planning becomes efficient, collaborative, and better. © 2025 IEEE.

关键词： Budget control

来源：评论

学校读者我要写书评

暂无评论

SmokeAway: An Early Smoke Detection IoT Architecture for Outdoors

SmokeAway: An Early Smoke Detection IoT Architecture for Out...

引用

2023 IEEE Global Conference on Artificial Intelligence and Internet of Things, GCAIoT 2023

作者： Alqarafi, Lama Aloufi, Norah Alluqmani, Rahaf Alrehaili, Shatha Alotaibi, Modhawi Taibah University College of Computer Science and Engineering Department of Computer Science Medina Saudi Arabia

ISBN: (纸本)9798350382020

The lack of available smoke detectors outdoors constitutes a significant factor contributing to the rapid spread of fires before effective containment can be achieved. The integration of outdoor smoke detectors helps tremendously with reducing fire and smoke hazards, consequently fostering a safe environment. With the emergence of smart cities and the revolutionary shift in relying on technologies such as the Internet of Things (IoT) and artificial intelligence (AI) tools, the current stage requires developing systems that integrate these tools and leverage them. To address that, this paper presents SmokeAway, a reliable system to promptly and accurately detect outdoor smoke. In order to do so, we propose a layered IoT architecture that relies on computer vision and a deep learning model. The efficacy of our model is assessed through two key performance metrics: accuracy and categorical cross-entropy loss. Compared to other models, our model has achieved notably low loss values, and accuracy rates of up to 96%, outperforming alternate models. © 2023 IEEE.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

Tokenizing Art Unpacking the Impact of Non-Fungible Tokens (NFTs) on Ownership Paradigms

Tokenizing Art Unpacking the Impact of Non-Fungible Tokens (...

引用

2024 International Conference on Computational Intelligence and Computing Applications, ICCICA 2024

作者： Jain, Atishay Joshi, Megha Rishiraj, Adarsh Kumar Kaur, Prabhjot Chandigarh University Department Of Computer Science And Engineering Mohali India

ISBN: (纸本)9798350306446

In the realm of the art of ownership, non-fungible Tokens (NFTs) have emerged as a revolutionary technological advancement, challenging conventional ideas of value and authenticity. This comprehensive review critically explores the intersection of NFTs and the art world, delving into their historical evolution, impact on the art market, and implications for artistic expression. By meticulously analyzing existing literature and case studies, this paper illuminates the transformative potential of NFTs, addressing the opportunities and challenges faced by artists, collectors, and stakeholders. The review focuses on key aspects such as legal and copyright considerations, technological constraints, and broader societal and environmental impacts associated with NFTs adoption in the art industry. Through a critical examination of the challenges and limitations of integrating NFTs into the art world, this review identifies areas for innovative solutions and future improvements. It emphasizes the importance of sustainable practice to reduce the environmental impact of NFT transactions and advocates for regularity frameworks to protect artist rights and foster inclusivity. Additionally, the paper underscores the advancements and innovations enabled by NFT adoption, highlighting their role in empowering artists and transforming the dynamics of art ownership and distribution. By suggesting future research directions and opportunities for collaboration, this study contributes to a nuanced understanding of the evolving landscape of art ownership in the digital age. © 2024 IEEE.

关键词： Environmental impact

来源：评论

学校读者我要写书评

暂无评论

Cloud Based Data Analytics for Business Intelligence 2

Cloud Based Data Analytics for Business Intelligence

引用

2nd International Conference on Advances in Computation, Communication and Information Technology, ICAICCIT 2024

作者： Talwandi, Navjot Singh Shah, Akhilesh Kumar Thakur, Rakesh Khare, Shanu Thakur, Payal Dept of Computer Science Engineering Chandigarh University Punjab India

ISBN: (纸本)9798331541217

This comprehensive review paper delves deeply into the realm of cloud-based data analytics platforms, underscoring their crucial role in propelling Business Intelligence (BI) forward during an era characterized by an unprecedented surge in data volumes. Faced with the deluge of data, organizations are increasingly relying on these platforms as indispensable tools to tap into the hidden potential of data and facilitate well-informed decision-making. The paper methodically traces the evolutionary path of cloud- based BI analytics platforms, shedding light on their diverse advantages and the substantial challenges they pose. It also elucidates the emerging trends in this dynamic landscape, outlining how these platforms are poised to reshape the future of BI. By conducting a thorough review of the existing body of literature, this paper imparts valuable insights into the defining attributes of these platforms, effective strategies for their adoption, and the critical factors that underpin their effectiveness. Additionally, it explores the pivotal role these platforms playing enhancing the flexibility, scalability, and cost-efficiency of BI processes. Moreover, in addition to furnishing an in-depth analysis, this review identifies significant gaps in current research and provides thoughtful recommendations for future avenues of exploration. These recommendations are intended to promote the continued growth and refinement of cloud-based data analytics platforms within the broader context of BI. © 2024 IEEE.

关键词： Business Intelligence Cloud Storage Data Analytics Stock

来源：评论

学校读者我要写书评

暂无评论

Skin Disease Detection System 13

Skin Disease Detection System

引用

13th International Conference on System Modeling and Advancement in Research Trends, SMART 2024

作者： Rehman, Fahad Aalim, Mohd Rahman, Nayeemur Chaudhary, Saurabhi Computer Science and Engineering Sharda University Greater Noida India

ISBN: (纸本)9798350380583

Given the fact that some of these diseases are difficult to establish diagnosis, including psoriasis, melanoma, eczema, seborrheic keratoses, and basal cell carcinoma, demand for better and more accurate diagnostic methods is always real. This paper will develop the application of CNN models based on the images sourced from the DermNet dataset to automatically classify the above five dermatological conditions. In addition, this dataset was processed with image scaling, normalization, and augmentation to ensure consistency and robustness of the method. The dimensionality was determined by feature extraction with the design of multiple convolutional layers of the CNN architecture. The model compilation was made by setting sparse categorical cross-entropy as the loss function combined with the Adam optimizer. The learned hyperparameters are the changes in the learning rate, batch size, and dropout rates. The metrics with Fl-score, precision, and recall set the means for judging the effectiveness of the model as all classes were said to score high accuracy for classification. This research, therefore, only serves to establish that AI models are relevant in the field of dermatology, though, there should be more research and clinical validation indeed. © 2024 IEEE.

关键词： Oncology

来源：评论

学校读者我要写书评

暂无评论

Enhancing Artificial Intelligence and Machine Learning Understanding Through EnVision: A Virtual Reality Approach 1

Enhancing Artificial Intelligence and Machine Learning Under...

引用

1st International Conference on Pioneering Developments in computer science and Digital Technologies, IC2SDT 2024

作者： Patra, Louise Kumari, Manisha Gopalapillai, Radhakrishnan Dept. of Computer Science and Engineering Karnataka Bengaluru India

ISBN: (纸本)9798350365016

The quality of teaching and learning process can be improved through innovative methods like use of virtual reality. We introduce EnVision, a groundbreaking Virtual Reality based approach to revolutionize Artificial Intelligence and Machine Learning (AI/ML) education. EnVision offers a unique approach where users build models through immersive learning experiences. Using interactive tutorials mixed with real-time data integration and visualization, users learn AI/ML concepts with ease. Through qualitative and quantitative methods, we evaluate EnVision's effectiveness and user satisfaction, contributing to the advancement of innovative educational gaming technologies. The evaluation results show that EnVision provides the AI/ML practitioners engaging learning experiences. © 2024 IEEE.

关键词： Virtual environments

来源：评论

学校读者我要写书评

暂无评论

Machine and deep learning approaches for heart disease risk assessment 4th

Machine and deep learning approaches for heart disease risk ...

引用

4th International Conference on Computational Methods in science and Technology, ICCMST 2024

作者： Sharma, Rahul Deepika Khan, Reshma Department of Computer Science and Engineering Chandigarh University Mohali Gharaun India

ISBN: (纸本)9781032911571

Heart disease is a major public health problem the world over, and accurate measures of risk assessment and appropriate classification of persons are very necessary in the early identification for treatment. This is one reason that the techniques of machine learning and deep learning find great application in the healthcare business, as they can analyze patterns of data and make predictions very effectively. The following research paper provides deep research on the techniques of machine learning (ML) and deep learning (DL) used to evaluate and categorize heart disease illness risk. It describes the different problems of the heart, the factors of risk, and models of ML and DL approaches. It also outlines what the current data sources in use are: digital health records, wearable technology, imaging, genetic data, and patient–reported outcomes. The review observes that, though challenges continue to be present in data collection, processing, and feature selection, it further has the ability to make the model free from bias and reliable. Finally, the authors state that such models might play a role in better patient outcomes and appropriate use of health resources, therefore allowing for focused medical strategies for the treatment of cardiac diseases. © 2025 the Author(s).

关键词： Electronic health record

来源：评论

学校读者我要写书评

暂无评论

A comparative study on data mining models for weather forecasting: A case study on Chittagong, Bangladesh

引用

Natural Hazards Research 2024年第2期4卷 295-303页

作者： Tahsin, Mohammad Sadman Abdullah, Shahriar Karim, Musaddiq Al Ahmed, Minhaz Uddin Tafannum, Faiza Ara, Mst Yeasmin Department of Computer Science and Engineering BRAC University Dhaka Bangladesh Department of Geography University of Alabama Tuscaloosa 35487 AL United States Department of Computer Science and Engineering American International University Bangladesh Dhaka Bangladesh Department of Computer Science and Engineering University of Texas at Arlington Texas 76019 TX United States

The primary focus of this study is to analyze and predict the patterns of this essential feature of the natural world. This study analyses and predicts the daily weather patterns of a specific urban area. This article utilizes weather data over 20 years to analyze the climate patterns of Chittagong city. A total of 12 distinct Data Mining models were employed to predict daily weather patterns. The algorithms can be categorized into three distinct types, namely rules-based, tree-based, and function-based. To evaluate the effectiveness of the models, various performance metrics were computed, including precision, recall, accuracy, F-measure, and the area under the receiver operating characteristic curve (ROC area). Based on the results obtained, it can be concluded that among the 12 algorithms evaluated, J48 exhibits the highest level of performance and accuracy. The J48 classifier demonstrated an accuracy of 82.30%, precision of 82.40%, recall of 82.20%, f-measure of 84.20%, and a ROC area of 97.8%. Furthermore, a comprehensive analysis of the confusion matrix for all twelve algorithms was conducted to facilitate further evaluation. © 2023 National Institute of Natural Hazards, Ministry of Emergency Management of China

关键词： Data mining Temperature Weather forecasting Weather prediction

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：