检索结果-内蒙古大学图书馆

IEEE International conference on Edge Computing (EDGE)

作者： Hadjer Benmeziane Kaoutar El Maghraoui IBM Research Europe Rüschlikon Switzerland IBM T. J. Watson Research Center Yorktown Heights NY USA

ISBN: (数字)9798350368499

ISBN: (纸本)9798350368505

Neural Architecture Search (NAS) is an incredibly indispensable technique in edge computing scenarios where computational resources are limited. However, the vastness of search spaces presents a formidable obstacle, hindering efficient exploration and identification of optimal architectures. In response, our study introduces an innovative paradigm that characterizes hardware and deep learning architectures, addressing the urgent demand for streamlined and swift architectural design *** approach pioneers the utilization of a fine-tuned Large language Model (LLM) to navigate the intricate terrain of architectural space. By harnessing the innate capabilities of LLMs, traditionally associated with natural language processing tasks, we extend their application domain to the realm of neural architecture design. empirical assessments conducted on benchmark datasets such as CIFAR-10 and ImageNet highlight the effectiveness of our fine-tuned LLM-based methodology. Notably, our findings reveal performance levels on par with stateof-the-art (SOTA) NAS techniques, achieving unparalleled speed and completing architectural searches within seconds.

关键词： Navigation Large language models Image edge detection Deep architecture Computer architecture natural language processing Hardware

来源：评论

学校读者我要写书评

暂无评论

Automatic Textual Normalization for Hate Speech Detection 1

引用

23rd International conference on Intelligent Systems Design and Applications, ISDA 2023

作者： Nguyen, Anh Thi-Hoang Nguyen, Dung Ha Nguyen, Nguyet Thi Ho, Khanh Thanh-Duy Nguyen, Kiet Van Faculty of Information Science and Engineering University of Information Technology Ho Chi Minh City Viet Nam Vietnam National University Ho Chi Minh City Viet Nam

ISBN: (数字)9783031647796

ISBN: (纸本)9783031647789

Social media data is a valuable resource for research, yet it contains a wide range of non-standard words (NSW). These irregularities hinder the effective operation of NLP tools. Current state-of-the-art methods for the Vietnamese language address this issue as a problem of lexical normalization, involving the creation of manual rules or the implementation of multi-staged deep learning frameworks, which necessitate extensive efforts to craft intricate rules. In contrast, our approach is straightforward, employing solely a sequence-to-sequence (Seq2Seq) model. In this research, we provide a dataset for textual normalization, comprising 2,181 human-annotated comments with an inter-annotator agreement of 0.9014. By leveraging the Seq2Seq model for textual normalization, our results reveal that the accuracy achieved falls slightly short of 70%. Nevertheless, textual normalization enhances the accuracy of the Hate Speech Detection (HSD) task by approximately 2%, demonstrating its potential to improve the performance of complex NLP tasks. Our dataset is accessible for research purposes (Github: https://***/AnhHoang0529/Small-LexNormViHSD). © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

关键词： natural language processing systems

来源：评论

学校读者我要写书评

暂无评论

Comparing Open Arabic Named Entity Recognition Tools 24

Comparing Open Arabic Named Entity Recognition Tools

引用

IEEE 24th International conference on Information Reuse and Integration for Data Science (IRI)

作者： Aldumaykhi, Abdullah Otai, Saad Alsudais, Abdulkareem Prince Sattam Bin Abdulaziz Univ Coll Comp Engn & Sci Al Kharj 11942 Saudi Arabia

ISBN: (纸本)9798350334586

The main objective of this paper is to compare and evaluate the performances of three open Arabic Named Entity Recognition (NER) tools: CAMeL, Hatmi, and Stanza. We collected a corpus consisting of 30 articles written in Modern Standard Arabic (MSA) and manually annotated all the entities of the person, organization, and location types at the article (document) level. Our results suggest a similarity between Stanza and Hatmi with the latter receiving the highest F1 score for the three entity types. However, CAMeL achieved the highest precision values for names of people and organizations. Following this, we implemented a "merge" method that combined the results from the three tools and a "vote" method that tagged named entities only when two of the three identified them as entities. Our results showed that merging achieved the highest overall F1 scores. Moreover, merging had the highest recall values while voting had the highest precision values for the three entity types. This indicates that merging is more suitable when recall is desired, while voting is optimal when precision is required. Finally, we collected a corpus of 21,635 articles related to COVID-19 and applied the merge and vote methods. Our analysis demonstrates the tradeoff between precision and recall for the two methods.

关键词： natural language processing Named Entity Recognition Platforms and Tools Software and Systems Reuse and Reusability

来源：评论

学校读者我要写书评

暂无评论

Leveraging AI-Based Large language Models to Strengthen English Vocabulary Skills in EFL Students

Leveraging AI-Based Large Language Models to Strengthen Engl...

引用

2024 International conference on Artificial Intelligence and Quantum Computation-Based Sensor Applications, ICAIQSA 2024

作者： Murugavel, S. Arunadevi, V. Chauhan, Rahul Punithaasree, K.S. Changala, Ravindra Bala, B Kiran Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology Department of English Avadi Chennai India Dr. Mahalingam College of Engineering and Technology Department of Science and Humanities Pollachi India Graphic Era Hill University Department of CSE Dehradun India Graphic Era Deemed to Be University Uttarakhand Dehradun India Department of English Erode India Guru Nanak Institutions Technical Campus Department of CSE Hyderabad India K.Ramakrishnan College of Engineering Department of AI & DS Trichy India

ISBN: (纸本)9798331517953

The concepts of AI and LLMs brought drastic changes to the sphere of educational technology, especially about the improvement of language teaching efficiency. Using elaborate computational algorithms and massive data, these models provide unprecedented potential for the identification and creation of related contexts hence are very efficient instruments in building up English vocabulary in students learning English as a Foreign language (EFL). In this paper, development of AI base vocabulary learning system with findings of advancement over the traditional and other modern methods in Python has been discussed. The indicated system relies on the GPT-4 model that was proved to offer natural language processing at a level higher than its predecessors thus differentiating the proposed vocabulary exercises based on context. The system also employs some of the modern practices like word embeddings (Word2Vec, GloVe), contextualized embeddings (BERT) and reinforcement learning to make it a dynamic learning system. The feasibility of the above-proposed AI-based system is assessed from the qualitative as well as quantitative analysis. Quantitative feedback is supported by qualitative feedback where results indicate higher student satisfaction while using the system. In general, the paper stands upon the idea that the proposed AI-based system could really improve the results of the study of English vocabulary with the benefits of this system towards traditional ones. This approach is indeed what can be described as the real progress in the field of educational technology, which provides a promising ground for the development of a strengthened approach to vocabulary instruction in EFL educational contexts. © 2024 IEEE.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

A Hybrid Deep Learning Approach for PM2.5 Concentration Prediction in Smart Environmental Monitoring

引用

Intelligent Automation & Soft Computing 2023年第6期36卷 3029-3041页

作者： Minh Thanh Vo Anh HVo Huong Bui Tuong Le Faculty of Information Technology HUTECH UniversityHo Chi Minh CityVietnam Natural Language Processing and Knowledge Discovery Laboratory Faculty of Information TechnologyTon Duc Thang UniversityHo Chi Minh CityVietnam Faculty of Computing Fundamentals FPT UniversityHo Chi Minh CityVietnam Laboratory for Artificial Intelligence Institute for Computational Science and Artificial IntelligenceVan Lang UniversityHo Chi Minh CityVietnam Faculty of Information Technology Van Lang UniversityHo Chi Minh CityVietnam

Nowadays,air pollution is a big environmental problem in develop-ing *** this problem,particulate matter 2.5(PM2.5)in the air is an air *** its concentration in the air is high in developing countries like Vietnam,it will harm everyone’s *** prediction of PM2.5 concentrations can help to make the correct decision in protecting the health of the *** study develops a hybrid deep learning approach named PM25-CBL model for PM2.5 concentration prediction in Ho Chi Minh City,***,this study analyzes the effects of variables on PM2.5 concentrations in Air Quality HCMC *** variables that affect the results will be selected for PM2.5 concentration ***,an efficient PM25-CBL model that integrates a convolutional neural network(CNN)andBidirectionalLongShort-TermMemory(Bi-LSTM)*** model consists of three following modules:CNN,Bi-LSTM,and Fully connected ***,this study conducts the experiment to compare the performance of our approach and several state-of-the-art deep learning models for time series prediction such as LSTM,Bi-LSTM,the combination of CNN and LSTM(CNN-LSTM),and *** empirical results confirm that PM25-CBL model outperforms other methods for Air Quality HCMC dataset in terms of several metrics including Mean Squared Error(MSE),Root Mean Squared Error(RMSE),Mean Absolute Error(MAE),and Mean Absolute Percentage Error(MAPE).

关键词： Time series prediction PM2.5 concentration prediction CNN Bi-LSTM network

来源：评论

学校读者我要写书评

暂无评论

A language-independent Metric for Measuring Text Simplification that does not Require a Parallel Corpus 35th

A Language-independent Metric for Measuring Text Simplificat...

引用

35th International Florida Artificial Intelligence Research Society conference, FLAIRS-35 2022

作者： Mucida, Lucas Oliveira, Alcione de P. Possi, Maurilio de A. Federal University of Vicosa Minas Gerais Vicosa Brazil

natural language processing encompasses several tasks, one of which is the automatic text simplification. Telling whether one text is simpler than another involves not only knowledge about the language being analyzed, but also a cultural knowledge of the target audience to which the text is being directed. Most of the current metrics used to measure text simplification are based on the use of parallel corpora, prepared by humans, which makes it difficult to apply the metrics in automatic text simplification in real time. In this paper, we present ISiM (Independent Simplification Metric), a metric that dismiss a parallel corpus, is simple, fast, language and human annotation independent, capable of quantifying the simplicity/complexity of a sentence, thus contributing improve automating text simplification. The results of the tests performed indicate that the proposed metric has the potential to be used to evaluate automatic methods of simplification. © 2021 by the authors. All rights reserved

关键词： natural language processing systems

来源：评论

学校读者我要写书评

暂无评论

Generative Adversarial Approach in natural language processing 25

Generative Adversarial Approach in Natural Language Processi...

引用

25th International conference on Soft Computing and Measurements, SCM 2022

作者： Karuna, Evgeniy N. Sokolov, Petr V. Gavrilic, Daria A. Saint Petersburg Electrotechnical University "lETI" Automatic Control Systems Department Saint Petersburg Russia Polytechnic University Peter the Great Saint Petersburg Saint Petersburg Russia

ISBN: (纸本)9781665496698

The use of a generative adversarial algorithm for training neural networks made it possible to make significant progress in solving the problem of generating images and audio data. Nevertheless, important problems remain in solving the tasks of generating discrete data sequences. Solving such problems will allow using generative-Adversarial learning to generate text data. This paper reflects a brief overview of modern research and achievements in the generation of text data using generative adversarial learning, lists a set of tasks that can be solved using this approach, describes possible problems and existing methods for solving existing problems, and also describes some suggestions for improving models. The structure and algorithm of the proposed system are described, the research results are presented. © 2022 IEEE.

关键词： natural language processing systems

来源：评论

学校读者我要写书评

暂无评论

Trade-offs between machine learning and deep learning for mental illness detection on social media

引用

SCIENTIFIC REPORTS 2025年第1期15卷 1-14页

作者： Ding, Zhanyi Wang, Zhongyan Zhang, Yeyubei Cao, Yuchen Liu, Yunchong Shen, Xiaorui Tian, Yexin Dai, Jianglai NYU Ctr Data Sci New York NY USA Univ Penn Sch Engn & Appl Sci Philadelphia PA USA Univ Calif Berkeley Dept EECS Berkeley CA 94720 USA Northeastern Univ Khoury Coll Comp Sci Boston MA USA Georgia Inst Technol Coll Comp Atlanta GA USA

Social media platforms provide valuable insights into mental health trends by capturing user-generated discussions on conditions such as depression, anxiety, and suicidal ideation. Machine learning (ML) and deep learning (DL) models have been increasingly applied to classify mental health conditions from textual data, but selecting the most effective model involves trade-offs in accuracy, interpretability, and computational efficiency. This study evaluates multiple ML models, including logistic regression, random forest, and LightGBM, alongside DL architectures such as ALBERT and Gated Recurrent Units (GRUs), for both binary and multi-class classification of mental health conditions. Our findings indicate that ML and DL models achieve comparable classification performance on medium-sized datasets, with ML models offering greater interpretability through variable importance scores, while DL models are more robust to complex linguistic patterns. Additionally, ML models require explicit feature engineering, whereas DL models learn hierarchical representations directly from text. Logistic regression provides the advantage of capturing both positive and negative associations between features and mental health conditions, whereas tree-based models prioritize decision-making power through split-based feature selection. This study offers empirical insights into the advantages and limitations of different modeling approaches and provides recommendations for selecting appropriate methods based on dataset size, interpretability needs, and computational constraints.

关键词： Machine learning Deep learning Mental health detection Social media natural language processing Model interpretability Feature importance

来源：评论

学校读者我要写书评

暂无评论

Efficient Nearest Neighbor Search in Large-Scale Text Data Using Optimized Locality Sensitive Hashing in Distributed Computing Environments

Efficient Nearest Neighbor Search in Large-Scale Text Data U...

引用

Advanced Computing Technologies (ICoACT), International conference on

作者： Shreya G. Tendulkar Pune University Pune India

ISBN: (数字)9798331509859

ISBN: (纸本)9798331509866

Efficient nearest neighbor search is crucial for applications such as web search, mobile browsing, and natural language processing (NLP), where rapid text retrieval from vast datasets is required. As data scales, traditional brute- force approaches become computationally expensive, necessitating optimized methods. This work introduces an enhanced Locality Sensitive Hashing (LSH) framework tailored for distributed computing environments, leveraging Hadoop for large-scale processing. By incorporating optimized LSH variants, our approach improves efficiency while maintaining high recall rates. Experimental evaluations demonstrate that our method surpasses conventional LSH implementations in retrieval accuracy while ensuring computational scalability, making it suitable for real-world applications requiring rapid text similarity assessments.

关键词： Social networking (online) Semantic search Scalability Heuristic algorithms Refining Nearest neighbor methods natural language processing Real-time systems Computational efficiency Web search

来源：评论

学校读者我要写书评

暂无评论

DeepDrugmiR: a two-stage deep learning method for inferring small molecules' regulatory effects on microRNA expression 24

DeepDrugmiR: a two-stage deep learning method for inferring ...

引用

15th conference on Bioinformatics Computational Biology and Health Informatics

作者： Huang, Yixian Wu, Huacong Cai, Ying Dong, Danhong Yu, Sicong Chen, Yigang Zhu, Zihao Lin, Yang-Chi-Dung Huang, Hsi-Yuan Huang, Hsien-Da Chinese Univ Hong Kong Sch Med Shenzhen Peoples R China Chinese Univ Hong Kong Sch Data Sci Shenzhen Peoples R China

ISBN: (纸本)9798400713026

MicroRNAs (miRNAs), vital regulators of gene expression and human health, are intimately associated with diseases upon dysregulation. Small molecules have emerged as promising therapeutic agents capable of modulating miRNA expression. However, laboratory-based identification of these regulatory effects is costly and time-consuming, highlighting the need for efficient and accurate computational models. Therefore, we propose DeepDrugmiR, a two-stage deep learning predictor that revolutionizes the prediction of small molecule-miRNA associations and their regulatory effects on miRNA expression. Leveraging novel language models tailored for chemical and DNA data, DeepDrugmiR encodes small molecules using double chemical language models and miRNAs using DNABERT-2, a nucleic acid language model. Its unique application of capsule networks, a next-generation neural architecture, further enhances prediction accuracy by capturing intricate relationships within small molecule structures and miRNA sequences. On benchmark datasets, DeepDrugmiR achieves state-of-the-art accuracy: 97% in identifying small molecule-miRNA associations and 92.7% (upregulation) & 92.4% (down-regulation) in predicting regulatory effects. This two-stage framework surpasses traditional single-stage methods in both information processing efficiency and noise resilience. A sorafenib case study validates DeepDrugmiR's usability, demonstrating consistency with miRNA profiling data. For ease of access, we have developed an intuitive web portal (https://***/similar to DeepDrugmiR/) to disseminate this innovative tool. By accelerating the understanding of drug-induced miRNA regulatory mechanisms, DeepDrugmiR propels the field of miRNA-targeted therapy forward.

关键词： small molecules regulatory effect microRNA expression language models capsule network web-based prediction tool miRNA-targeted therapy

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：