检索结果-内蒙古大学图书馆

39th Annual AAAI Conference on Artificial Intelligence, AAAI 2025

作者： Gong, Ze Kumar, Akshat Varakantham, Pradeep School of Computing and Information Systems Singapore Management University Singapore

ISBN: (纸本)157735897X

Offline safe reinforcement learning (RL) has emerged as a promising approach for learning safe behaviors without engaging in risky online interactions with the environment. Most existing methods in offline safe RL rely on cost constraints at each time step (derived from global cost constraints) and this can result in either overly conservative policies or violation of safety constraints. In this paper, we propose to learn a policy that generates desirable trajectories and avoids undesirable trajectories. To be specific, we first partition the pre-collected dataset of state-action trajectories into desirable and undesirable subsets. Intuitively, the desirable set contains high reward and safe trajectories, and undesirable set contains unsafe trajectories and low-reward safe trajectories. Second, we learn a policy that generates desirable trajectories and avoids undesirable trajectories, where (un)desirability scores are provided by a classifier learnt from the dataset of desirable and undesirable trajectories. This approach bypasses the computational complexity and stability issues of a min-max objective that is employed in existing methods. Theoretically, we also show our approach’s strong connections to existing learning paradigms involving human feedback. Finally, we extensively evaluate our method using the DSRL benchmark for offline safe RL. Empirically, our method outperforms competitive baselines, achieving higher rewards and better constraint satisfaction across a wide variety of benchmark tasks. Copyright © 2025, Association for the Advancement of Artificial Intelligence (***). All rights reserved.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Generating bilingual example sentences with large language models as lexicography assistants 22nd

Generating bilingual example sentences with large language m...

引用

22nd Annual Workshop of the Australasian Language Technology Association, ALTA 2024

作者： Merx, Raphael Vylomova, Ekaterina Kurniawan, Kemal School of Computing and Information Systems The University of Melbourne Australia

We present a study of LLMs’ performance in generating and rating example sentences for bilingual dictionaries across languages with varying resource levels: French (high-resource), Indonesian (mid-resource), and Tetun (low-resource), with English as the target language. We evaluate the quality of LLM-generated examples against the GDEX (Good Dictionary EXample) criteria: typicality, informativeness, and intelligibility (Kilgarriff et al., 2008). Our findings reveal that while LLMs can generate reasonably good dictionary examples, their performance degrades significantly for lower-resourced languages. We also observe high variability in human preferences for example quality, reflected in low inter-annotator agreement rates. To address this, we demonstrate that in-context learning can successfully align LLMs with individual annotator preferences. Additionally, we explore the use of pre-trained language models for automated rating of examples, finding that sentence perplexity serves as a good proxy for "typicality" and "intelligibility" in higher-resourced languages. Our study also contributes a novel dataset of 600 ratings for LLM-generated sentence pairs, and provides insights into the potential of LLMs in reducing the cost of lexicographic work, particularly for low-resource languages. © 2024 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

ADA-Gen: Iterative and Incremental Generation of Full-Stack Apps for Learning Agile/DevOps Software Development Practices 17

ADA-Gen: Iterative and Incremental Generation of Full-Stack ...

引用

17th International Conference on Computer Supported Education, CSEDU 2025

作者： Duong, Ta Nguyen Binh School of Computing and Information Systems Singapore Management University Singapore

ISBN: (纸本)9789897587467

To learn Agile/DevOps practices effectively, students need to apply them in an actual software development project. This is challenging if students are mostly from non-computing backgrounds and they do not have time in the curriculum to learn programming and related tools. Therefore, it is important to help students who do not possess programming foundations to develop fully functional software during the process of learning Agile/DevOps concepts. We noted that existing low-code/no-code app development platforms have not been designed to teach Agile/DevOps practices. On the other hand, recent AI-based tools for code generation such as GitHub Copilot have been built mainly for programmers. In this work, we designed and implemented ADA-Gen (Agile/DevOps App Generator), a teaching tool leveraging large language models (LLMs) to generate full-stack web apps following an iterative and incremental development methodology widely practiced in Agile/DevOps circles. ADA-Gen is integrated with Jira, one of the most popular platforms for software project management, so students can use it right away without much further setup. This approach allows non-computing students to experience the complete full-stack software development life cycle. We have conducted extensive evaluations of ADA-Gen using various realistic project scenarios. The evaluations demonstrated the capabilities of ADA-Gen in full-stack web app generation, and in providing plenty of learning opportunities for students to appreciate key Agile/DevOps practices. Copyright © 2025 by SCITEPRESS - Science and Technology Publications, Lda.

关键词： Software design

来源：评论

学校读者我要写书评

暂无评论

User Willingness to Engage with AI-Generated Influencer Content

User Willingness to Engage with AI-Generated Influencer Cont...

引用

2025 CHI Conference on Human Factors in computing systems, CHI EA 2025

作者： Sew, Cherie Schömbs, Sarah Kelly, Ryan M. Goncalves, Jorge School of Computing and Information Systems University of Melbourne Melbourne Australia School of Computing and Information Systems The University of Melbourne MelbourneVIC Australia School of Computing Technologies RMIT University MelbourneVIC Australia

ISBN: (纸本)9798400713958

In this study, we investigate the reasons behind social media users’ willingness or reluctance to engage with AI-generated influencers, an increasingly prevalent presence on social media platforms. We conducted a user study with 120 participants to gather their opinions and perceptions of AI-generated influencers after they watched four short-form videos featuring different AI-generated influencers with varying degrees of human-likeness. We grouped our findings into four high-level categories: positive perceptions, influence of message content, reservations about AI-generated influencers, and shifting views and future expectations. From these categories, we present ten sub-categories as potential drivers for willingness or reluctance in user engagement. Additionally, we discuss the implications, ethics, and risks of utilising AI-generated influencers. © 2025 Copyright held by the owner/author(s).

关键词： Tweets

来源：评论

学校读者我要写书评

暂无评论

Deep learning in crowd counting:A survey

引用

CAAI Transactions on Intelligence Technology 2024年第5期9卷 1043-1077页

作者： Lijia Deng Qinghua Zhou Shuihua Wang Juan Manuel Górriz Yudong Zhang School of Computing and Mathematical Sciences University of LeicesterLeicesterUK Department of Information Systems Faculty of Computing and Information TechnologyKing Abdulaziz UniversityJeddahSaudi Arabia Department of Signal Theory Networking and CommunicationsUniversity of GranadaGranadaSpain

Counting high-density objects quickly and accurately is a popular area of *** counting has significant social and economic value and is a major focus in artificial *** many advancements in this field,many of them are not widely known,especially in terms of research *** authors proposed a three-tier standardised dataset taxonomy(TSDT).The Taxonomy divides datasets into small-scale,large-scale and hyper-scale,according to different application *** theory can help researchers make more efficient use of datasets and improve the performance of AI algorithms in specific ***,the authors proposed a new evaluation index for the clarity of the dataset:average pixel occupied by each object(APO).This new evaluation index is more suitable for evaluating the clarity of the dataset in the object counting task than the image ***,the authors classified the crowd counting methods from a data-driven perspective:multi-scale networks,single-column networks,multi-column networks,multi-task networks,attention networks and weak-supervised networks and introduced the classic crowd counting methods of each *** authors classified the existing 36 datasets according to the theory of three-tier standardised dataset taxonomy and discussed and evaluated these *** authors evaluated the performance of more than 100 methods in the past five years on different levels of popular ***,progress in research on small-scale datasets has slowed *** are few new datasets and algorithms on small-scale *** studies focused on large or hyper-scale datasets appear to be reaching a saturation *** combined use of multiple approaches began to be a major research *** authors discussed the theoretical and practical challenges of crowd counting from the perspective of data,algorithms and computing *** field of crowd counting is moving towards combining multiple methods and requ

关键词： artificial intelligence computer vision image analysis image processing

来源：评论

学校读者我要写书评

暂无评论

Meaningful Engagement, Ethical Care, and Design Opportunities: An Ethnographic Study on Social Activities in Long-term Care 25

Meaningful Engagement, Ethical Care, and Design Opportunitie...

引用

2025 CHI Conference on Human Factors in computing systems, CHI 2025

作者： Yuan, Shuai Coghlan, Simon Lederman, Reeva Waycott, Jenny School of Computing and Information Systems The University of Melbourne MelbourneVIC Australia

ISBN: (纸本)9798400713941

Social activities in long-term care homes help promote residents' wellbeing, but their effectiveness depends on residents' engagement. To identify design opportunities for promoting meaningful engagement, we conducted an ethnographic study on organised activities in an Australian aged care home. We observed staff fostered engagement by initiating conversations, weaving residents' backgrounds into interactions, and adapting activities to residents' varying abilities. However, challenges included new staff members' unfamiliarity with residents, multi-tasking, and insufficient support to engage excluded residents. Using a care ethics lens that includes relational, situated and empathetic features of care, we show that meaningful engagement is shaped by the ethical care practices embedded in staff-resident interactions and highlight opportunities for technologies to mitigate barriers hindering staff from providing ethical care in existing activities. These opportunities include: collecting and recording residents' interests, providing conversation prompts, enhancing activity inclusiveness, and reducing language and cultural barriers. © 2025 Copyright held by the owner/author(s).

关键词： Ethical technology

来源：评论

学校读者我要写书评

暂无评论

Digital transformation and the future of work: closing the digital skills gap

引用

Development and Learning in Organizations 2024年第3期39卷 14-17页

作者： Hoe, Siu Loon School of Computing and Information Systems Singapore Management University Singapore Singapore

Purpose: The purpose of this article is to discuss the near future digital technology landscape and propose several specific in-demand digital skills for organizations and individuals in the next few years. Design/methodology/approach: This article reviews some recent publications from representative inter-governmental, governmental, non-governmental, and commercial organizations on the rise of digital technologies and corresponding growth in digital jobs. Within this context, several specific in-demand skills are proposed by the author who has written a book on the topic of digital transformation. Findings: Rapid advancements in digital technologies continue to shape organizational practices and the future of work. To take advantage of emerging digital technologies and better compete in the marketplace, organizations and individuals need to upskill and reskill. Practical implications: The main purpose of this article is to suggest a list of specific digital skills for executive management and human resources professionals to help prepare employees for digital jobs in the next few years. Corporate policy makers could introduce and promote these digital skills as part of the organizations’ training and development curriculum. In the area of talent management, organizations could also formulate hiring policies and assessment mechanisms to attract, develop, and retain employees with the requisite digital skills. For information technology and business professionals seeking to upskill and reskill, they could acquire these digital skills to enhance their current roles and prepare for the future digital workplace environment. Originality/value: This article contributes to a better understanding of the digital technology landscape and digital skills. A list of specific in-demand skills is proposed to help organizations and individuals close the digital skills gap. © 2024, Emerald Publishing Limited.

关键词： Competency Development Digital age Jobs Talent management Training

来源：评论

学校读者我要写书评

暂无评论

Priority Detector and Classifier Techniques Based on ML for the IoMT

引用

Computers, Materials & Continua 2023年第8期76卷 1853-1870页

作者： Rayan A.Alsemmeari Mohamed Yehia Dahab Badraddin Alturki Abdulaziz A.Alsulami Department of Information Technology Faculty of Computing and Information TechnologyKing Abdulaziz UniversityJeddah21589Saudi Arabia Department of Computer Science Faculty of Computing and Information TechnologyKing Abdulaziz UniversityJeddah21589Saudi Arabia Department of Information Systems Faculty of Computing and Information TechnologyKing Abdulaziz UniversityJeddah21589Saudi Arabia

Emerging telemedicine trends,such as the Internet of Medical Things(IoMT),facilitate regular and efficient interactions between medical devices and computing *** importance of IoMT comes from the need to continuously monitor patients’health conditions in real-time during normal daily activities,which is realized with the help of various wearable devices and *** major health problem is workplace stress,which can lead to cardiovascular disease or psychiatric ***,real-time monitoring of employees’stress in the workplace is *** levels and the source of stress could be detected early in the fog layer so that the negative consequences can be mitigated ***,overwhelming the fog layer with extensive data will increase the load on fog nodes,leading to computational *** study aims to reduce fog computation by proposing machine learning(ML)models with two *** first phase of theMLmodel assesses the priority of the situation based on the stress *** the second phase,a classifier determines the cause of stress,which was either interruptions or time pressure while completing a *** approach reduced the computation cost for the fog node,as only high-priority records were transferred to the ***-priority records were forwarded to the *** MLapproaches were compared in terms of accuracy and prediction speed:Knearest neighbors(KNN),a support vector machine(SVM),a bagged tree(BT),and an artificial neural network(ANN).In our experiments,ANN performed best in both phases because it scored an F1 score of 99.97% and had the highest prediction speed compared with KNN,SVM,and BT.

关键词： Machine learning priority detector Internet of Medical Things IoMT fog computing cloud computing

来源：评论

学校读者我要写书评

暂无评论

Designing Embodied Agents for Impactful Human-Centred Risk Communication

Designing Embodied Agents for Impactful Human-Centred Risk C...

引用

2025 CHI Conference on Human Factors in computing systems, CHI EA 2025

作者： Schömbs, Sarah School of Computing and Information Systems The University of Melbourne MelbourneVIC Australia

ISBN: (纸本)9798400713958

Embodied agents are increasingly accessible to users and deployed in critical domains, which creates an urgent need for intuitive, human-centred risk communication to ensure users understand potential risks and infer appropriate action. My research explores embodied agents as novel interfaces for risk communication, investigating how these agents can intuitively convey risk information and facilitate risk alignment as an integral part of risk communication. I examine how using embodied agents to communicate risk impacts users’ risk perception, decision-making, and understandability. Building on this work, my proposed research focuses on how agents can further support risk alignment by allowing users to specify and adjust their risk preferences through risk alignment interfaces. My dissertation contributes empirical insights, implications, and design considerations for human-centred risk communication in human-agent interaction and contributes to HCI by offering important insights for designing agents that prioritise risk comprehension and appropriate risk responses in complex environments. © 2025 Copyright held by the owner/author(s).

关键词： Risk assessment

来源：评论

学校读者我要写书评

暂无评论

Input-Based Ensemble-Learning Method for Dynamic Memory Configuration of Serverless computing Functions 17

Input-Based Ensemble-Learning Method for Dynamic Memory Conf...

引用

17th IEEE/ACM International Conference on Utility and Cloud computing, UCC 2024

作者： Agarwal, Siddharth Rodriguez, Maria A. Buyya, Rajkumar Laboratory School of Computing and Information Systems The University of Melbourne Australia

ISBN: (纸本)9798350367201

In today’s Function-as-a-Service offerings, a programmer is usually responsible for configuring function memory for its successful execution, which allocates proportional function resources such as CPU and network. However, right-sizing the function memory force developers to speculate performance and make ad-hoc configuration decisions. Recent research has highlighted that a function’s input characteristics, such as input size, type and number of inputs, significantly impact its resource demand, run-time performance and costs with fluctuating workloads. This correlation further makes memory configuration a non-trivial task. On that account, an input-aware function memory allocator not only improves developer productivity by completely hiding resource-related decisions but also drives an opportunity to reduce resource wastage and offer a finer-grained cost-optimised pricing scheme. Therefore, we present MemFigLess, a serverless solution that estimates the memory requirement of a serverless function with input-awareness. The framework executes function profiling in an offline stage and trains a multi-output Random Forest Regression model on the collected metrics to invoke input-aware optimal configurations. We evaluate our work with the state-of-the-art approaches on AWS Lambda service to find that MemFigLess is able to capture the input-aware resource relationships and allocate upto 82% less resources and save up to 87% run-time costs. © 2024 IEEE.

关键词： Resource allocation

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：