Deep Reinforcement Learning Empowered Edge Collaborative Caching Scheme for Internet of Vehicles

With the development of internet of vehicles,the traditional centralized content caching mode transmits content through the core network,which causes a large delay and cannot meet the demands for delay-sensitive services.To solve these problems,on basis of vehicle caching network,we propose an edge colla-borative caching scheme.Road side unit(RSU)and mobile edge computing(MEC)are used to collect vehicle information,predict and cache popular content,thereby provide low-latency content delivery services.However,the storage capa-city of a single RSU severely limits the edge caching performance and cannot handle intensive content requests at the same time.Through content sharing,col-laborative caching can relieve the storage burden on caching servers.Therefore,we integrate RSU and collaborative caching to build a MEC-assisted vehicle edge collaborative caching(MVECC)scheme,so as to realize the collaborative caching among cloud,edge and vehicle.MVECC uses deep reinforcement learning to pre-dict what needs to be cached on RSU,which enables RSUs to cache more popular content.In addition,MVECC also introduces a mobility-aware caching replace-ment scheme at the edge network to reduce redundant cache and improving cache efficiency,which allows RSU to dynamically replace the cached content in response to the mobility of vehicles.The simulation results show that the pro-posed MVECC scheme can improve cache performance in terms of energy cost and content hit rate.

Research on Collaborative Learning in Network Learning Environment

This paper summarizes the basic content of network curriculum design based on online learning mode and the basic flow, as well as network course should have the factors that suitable of the mode and attention matters in the design collaboration mode of network course. Based on this, other researchers and practitioners can conveniently and effectively design network course based on the cooperation mode. Through the analysis of the network curriculum development and the actual case, verify advantage of collaborative online learning mode.

A New Pattern to Research the Learning:From Phenomenography and Constructivism Perspective——The Case of Guangzhou Academy of Fine Arts

This essay is trying to explore how the arts students experience the teaching and learning context in the English class of the col lege from the phenomenography perspective.The qualitative research methodology of phenomenography has traditionally required a man ual sorting and analysis of interview data.To study the teaching and learning context,the qualitative research method will be applied,and the data collection will base on the face-to-face interview.There are 8 sophomores were interviewed after they had one year study in col lege.The research findings reveal that most student in interview gradually get used to college English class,but never think about changing their learning approach even they study one year in a totally different teaching and learning environment.

Tackling Faceless Killers: Toxic Comment Detection to Maintain a Healthy Internet Environment

According to BBC News,online hate speech increased by 20%during the COVID-19 pandemic.Hate speech from anonymous users can result in psychological harm,including depression and trauma,and can even lead to suicide.Malicious online comments are increasingly becoming a social and cultural problem.It is therefore critical to detect such comments at the national level and detect malicious users at the corporate level.To achieve a healthy and safe Internet environment,studies should focus on institutional and technical topics.The detection of toxic comments can create a safe online environment.In this study,to detect malicious comments,we used approxi-mately 9,400 examples of hate speech from a Korean corpus of entertainment news comments.We developed toxic comment classification models using supervised learning algorithms,including decision trees,random forest,a support vector machine,and K-nearest neighbors.The proposed model uses random forests to classify toxic words,achieving an F1-score of 0.94.We analyzed the trained model using the permutation feature importance,which is an explanatory machine learning method.Our experimental results confirmed that the toxic comment classifier properly classified hate words used in Korea.Using this research methodology,the proposed method can create a healthy Internet environment by detecting malicious comments written in Korean.

A virtual learning environment of the Chinese University of Hong Kong

This paper introduces a scalable virtual learning environment of the ChineseUniversity of Hong Kong;an explicitly geographical, immersive, and sharable 3Dlearning space with comprehensive social elements. It is characterized by multiuser collaborative modeling, group learning approaches of geo-collaboration,social space-oriented hierarchical avatars, and knowledge exchanging and sharingbased on virtual geographic experiments. Applications for the purpose of publiceducation and virtual geographic experiment, and indicated future works provethe possibility to offer a greater opportunity to foster interdisciplinary collaborations, revitalize teaching patterns and learning contents, improve learners’cognitive abilities to solve problems, and enhance their understanding of scientificconcepts and processes.

Collaborative Virtual Learning in Education in STEM Education

There has been a recent explosion of interest from academics across a wide range of disciplines in the use of Multi-User Virtual Environments for education,driven by the success of platforms such as Internet Communication Technology learning skills in higher education.While digital virtual worlds are used in the 21st century learning,advances in the capabilities and the spread of technology have fed a recent boom in interest in massively multi-user 3D virtual worlds for entertainment,and this in turn has led to a surge of interest in their educational applications.As these platforms are used more often as environments for teaching and learning,there is increased need to integrate them with other institutional systems,Web-based Virtual Learning Environments(VLE)in particular.In this paper,we briefly review the use of virtual worlds for education,from informal learning to formal instruction,and consider what is required to turn a virtual world from a Multi-User Virtual Environment into a fully fledged 3D Virtual Learning Environment(VLE).In this we focus on the development of Moodle—a system which integrates the popular 3D virtual world of Second Life with the open-source Virtual Learning Environment.

Hybrid Deep Learning Enabled Air Pollution Monitoring in ITS Environment

Intelligent Transportation Systems(ITS)have become a vital part in improving human lives and modern economy.It aims at enhancing road safety and environmental quality.There is a tremendous increase observed in the number of vehicles in recent years,owing to increasing population.Each vehicle has its own individual emission rate;however,the issue arises when the emission rate crosses a standard value.Owing to the technological advances made in Artificial Intelligence(AI)techniques,it is easy to leverage it to develop prediction approaches so as to monitor and control air pollution.The current research paper presents Oppositional Shark Shell Optimization with Hybrid Deep Learning Model for Air Pollution Monitoring(OSSOHDLAPM)in ITS environment.The proposed OSSO-HDLAPM technique includes a set of sensors embedded in vehicles to measure the level of pollutants.In addition,hybridized Convolution Neural Network with Long Short-Term Memory(HCNN-LSTM)model is used to predict pollutant level based on the data attained earlier by the sensors.In HCNN-LSTM model,the hyperparameters are selected and optimized using OSSO algorithm.In order to validate the performance of the proposed OSSO-HDLAPM technique,a series of experiments was conducted and the obtained results showcase the superior performance of OSSO-HDLAPM technique under different evaluation parameters.

Intrusion Detection System for Smart Industrial Environments with Ensemble Feature Selection and Deep Convolutional Neural Networks

Smart Industrial environments use the Industrial Internet of Things(IIoT)for their routine operations and transform their industrial operations with intelligent and driven approaches.However,IIoT devices are vulnerable to cyber threats and exploits due to their connectivity with the internet.Traditional signature-based IDS are effective in detecting known attacks,but they are unable to detect unknown emerging attacks.Therefore,there is the need for an IDS which can learn from data and detect new threats.Ensemble Machine Learning(ML)and individual Deep Learning(DL)based IDS have been developed,and these individual models achieved low accuracy;however,their performance can be improved with the ensemble stacking technique.In this paper,we have proposed a Deep Stacked Neural Network(DSNN)based IDS,which consists of two stacked Convolutional Neural Network(CNN)models as base learners and Extreme Gradient Boosting(XGB)as the meta learner.The proposed DSNN model was trained and evaluated with the next-generation dataset,TON_IoT.Several pre-processing techniques were applied to prepare a dataset for the model,including ensemble feature selection and the SMOTE technique.Accuracy,precision,recall,F1-score,and false positive rates were used to evaluate the performance of the proposed ensemble model.Our experimental results showed that the accuracy for binary classification is 99.61%,which is better than in the baseline individual DL and ML models.In addition,the model proposed for IDS has been compared with similar models.The proposed DSNN achieved better performance metrics than the other models.The proposed DSNN model will be used to develop enhanced IDS for threat mitigation in smart industrial environments.

An Empirical Study of Legal English Translation Teaching Through Wechat-based Mobile Learning Platform

This paper illustrates the functions of smartphone-based teaching using the theory of constructivism,and puts forward anew learning strategy to replace traditional cram-teaching methods.We examines the new paradigm in the formation of translationcompetence within the legal discourse.It aims to promote the autonomous learning,monitor the students’participation,facilitatestudents’communication and provide well-structured materials to transform traditional classroom learning into mobile phonelearning,to maximize students’initiative and enthusiasm,as well as help students engage in,interpret,and negotiate the complexi-ties that surround them.The findings of this study have been summarized into a few generalizations for possible directions for trans-lation research and they provide a better understanding of Chinese students’translation competence within a legal English contextand contribute to the translation skill development.Series on existing research on translation competence development in classroomteaching contexts for empirical guidance,as an essential component of ESP curriculum based on authentic data and analyzedthrough online framework specifically designed for legal discourse.

IoTFLiP： IoT-based flipped learning platform for medical education

Case-Based Learning （CBL） has become an effective pedagogy for student-centered learning in medical education, which is founded on persistent patient cases. Flippped learning and Internet of Things （IoTs） concepts have gained significant attention in recent years. Using these concepts in conjunction with CBL can improve learning ability by providing real evolutionary medical eases. It also enables students to build confidence in their decision making, and efficiently enhances teamwork in the learning environment. We propose an IoT-based Flip Learning Platform, called IoTFLiP, where an IoT infrastrneture is exploited to support flipped case-based learning in a cloud environment with state of the art security and privacy measures for personalized medical data. It also provides support for application delivery in private, public, and hybrid approaches. The proposed platform is an extension of our Interactive Case-Based Flipped Learning Tool （ICBFLT）, which has been developed based on current CBL practices. ICBFLT formulates summaries of CBL cases through synergy between students＇ and medical expert knowledge. The low cost and reduced size of sensor device, support of IoTs, and recent flipped learning advancements can enhance medical students＇ academic and practical experiences. In order to demonstrate a working scenario for the proposed IoTFLiP platform, real-time data from IoTs gadgets is collected to generate a real-world case for a medical student using ICBFLT.

基于Internet的协作学习环境系统的设计与实现

建构主义学习理论强调协作学习对意义建构的重要作用,同时重视学习环境的设计,认为学习者在一定的环境中利用各种工具和信息资源完成意义的建构。Internet以诸多优势为开展协作学习在提供了有效的环境。该文分析了基于Internet的协作学习进程和环境系统目标,提出了一个基于Internet的协作学习环境系统的设计模型,并给出其实现方法。作为应用,作者简要描述了《摄影构图》在该系统上学习的过程。

基于Internet的小组协作学习平台比较研究

随着互联网的普及,出现了许多支持网络环境下的小组协作学习平台,对基于E-mail、Blog、BBS、视频会议系统、Wiki、QQ群等六种常用的小组学习平台进行对比分析和归纳总结,为教师在教学中选择小组协作学习平台提供参考依据.

基于Intranet/Internet的协作学习系统设计

基于Intranet/Internet的协作学习系统是一体化的有效学习系统。本文从协作学习理论依据及其模式;基于Intranet/Internet的学习系统构成入手,探讨如何进行基于Intranet/Internet的协作学习系统设计。

虚拟Internet教学环境

利用局域网建立虚拟Internet教学环境，解决校园网传输速度慢、费用高、教学效果差的普遍问题，改善Internet教学环境。

智慧学习环境中的人机协同设计

作为教育数字化转型的首要任务,智慧学习环境建设过分强调技术之于教育的能力,而忽略教育主体的价值与地位,涌现出场景割裂、数据孤岛等问题。人机协同旨在充分发挥人与机器的优势,弥补彼此的劣势,成为指导智慧学习环境创设与优化的最优解。研究将人机协同视为智慧学习环境设计的基线思维,构建了由数据模型层、技术支撑层和场景应用层三个层级,包含场景、数据、模型、资源、工具与服务等六个要素的智慧学习环境概念模型。基于普瑞斯的人机功能分配决策矩阵理论,提出了AI讲师、执行型AI+人类助手、伙伴型AI+人类同侪、助教型AI+人类教练、人类导师等五种人机协同模式。在此基础上,研究制定了智慧学习环境各层级的设计原则,分析了数据模型层的决策协同设计、技术支撑层的交互协同设计和场景应用层的流程协同设计,讨论了人机协同模式中人机互信和价值对齐的建构策略,以期指导智慧学习环境中的人机协同设计。

基于异步深度强化学习的车联网协作卸载策略

随着车联网(IoV)的快速发展,智能网联汽车产生了大量延迟敏感型和计算密集型任务,有限的车辆计算资源以及传统的云服务模式无法满足车载用户的需求,移动边缘计算(MEC)为解决海量数据的任务卸载提供了一种有效范式。但在考虑多任务、多用户场景时,由于车辆位置、任务种类以及车辆密度的实时性和动态变化,IoV中任务卸载场景复杂度较高,卸载过程中容易出现边缘资源分配不均衡、通信成本开销过大、算法收敛慢等问题。为解决以上问题,重点研究了IoV中多任务、多用户移动场景中的多边缘服务器协同任务卸载策略。首先,提出一种多边缘协同处理的三层异构网络模型,针对IoV中不断变化的环境,引入动态协作簇,将卸载问题转化为时延和能耗的联合优化问题;其次,将问题分为卸载决策和资源分配两个子问题,其中资源分配问题又拆分为面向边缘服务器和传输带宽的资源分配,并基于凸优化理论求解。为了寻求最优卸载决策集,提出一种能在协作簇中处理连续问题的多边缘协作深度确定性策略梯度(MC-DDPG)算法,并在此基础上设计了一种异步多边缘协作深度确定性策略梯度(AMCDDPG)算法,通过将协作簇中的训练参数异步上传至云端进行全局更新,再将更新结果返回每个协作簇中提高收敛速度。仿真结果显示,AMC-DDPG算法较DDPG算法至少提高了30%的收敛速度,且在奖励和总成本等方面也取得了较好的效果。

HMFuzzer:一种基于人机协同的物联网设备固件漏洞挖掘方案

模糊测试是一种针对物联网设备固件漏洞挖掘的主流方法,能够先攻击者一步发现安全威胁,提升物联网设备的安全性.但是目前大部分的模糊测试技术关注于如何自动化地实现漏洞挖掘,忽略了专家经验对于设备固件漏洞挖掘工作的优势.本文提出一种基于人机协同的物联网设备固件漏洞挖掘方案HMFuzzer,设计了基于设备固件前后端交互的设备固件关键信息提取方法,通过模拟设备固件、设备管理界面以及用户三方交互模式获取固件潜在的关键信息,并通过二进制文件定位和函数分析技术解析出固件关键函数.此外,HMFuzzer通过在模糊测试的预处理、测试和结果分析阶段引入专家经验,利用上一阶段获取的关键信息,结合强化学习算法,优化种子变异和模糊测试流程,显著提升了模糊测试的覆盖率、效率以及漏洞挖掘能力.实验结果表明,相比于现有的固件漏洞挖掘方法,HMFuzzer的漏洞识别成功率能提高10%以上,具备更强的漏洞检测能力.特别是,针对真实厂商的物联网设备测试,HMFuzzer发现了多个0-day漏洞,其中已获得4个CVE/CNVD高危漏洞.

面向6G物联网设备协同的区块链动态分片

随着物联网规模化应用的不断落地,海量设备协同工作,产生了大量高价值数据。这些数据若得不到有效的安全保障,就容易遭受数据滥用、隐私泄露以及数据篡改等威胁。因此,去中心化、不可篡改、安全的区块链分片网络逐步取代传统的集中式网络,被应用到该场景中。然而,区块链分片网络受限于复杂环境以及高比例跨片协同事务。针对上述问题,提出一种面向6G物联网设备协同的区块链动态分片优化方案。设计分片系统架构,建立整个系统的吞吐量模型、安全模型以及时延模型。在此基础上,提出两阶段分片优化策略。第一阶段采用信誉分级分片策略筛选节点,第二阶段采用基于深度强化学习算法的动态分片策略,降低跨片协同事务比例,决策分片数量。两阶段的设计目的是在保证安全的情况下最大程度地提升整个系统的吞吐量。实验结果表明,在面向物联网设备协同的区块链分片场景下,相较于传统的基于单一的信誉分级分片策略、均匀分片策略或者随机分片策略的方案,所提方案在保证安全性的情况下,平均每轮减少50%以上的跨片协同事务比例,有效地提升了系统的吞吐量。

基于区块链的工业物联网隐私保护协作学习系统

为了在保护数据隐私的前提下,充分利用异构的工业物联网节点数据训练高精度模型,提出了一种基于区块链的隐私保护两阶段协作学习系统。首先,使用分组联邦学习框架,根据参与节点的算力将其划分为不同组,每组通过联邦学习训练一个适合其算力的全局模型;其次,引入分割学习,使节点能够与移动边缘计算服务器协作训练更大规模的模型,并采用差分隐私技术进一步保护数据隐私,将训练好的模型存储在区块链上,通过区块链的共识算法进一步防止恶意节点的攻击,保护模型安全;最后,为了结合多个异构全局模型的优点并进一步提高模型精度,使用每个全局模型的特征提取器从用户数据中提取特征,并将这些特征用作训练集训练更高精度的复杂模型。实验结果表明,该系统在Fashion-MNIST和CIFAR-10数据集上的性能优于传统联邦学习的性能,能够应用于工业物联网场景中以获得高精度模型。

智能协作学习环境中学习者的提问能力对认知过程的影响研究

人工智能技术的快速发展推动了智能协作学习环境的兴起。在此背景下,学习者的提问能力尤为关键,它能够直接增强学习者与智能系统的互动,提升学习体验和效果。然而,目前鲜有研究关注学习者的提问能力对其认知过程的影响。文章旨在探索智能协作学习环境中,学习者的提问能力如何影响他们的认知过程。通过对学习者提问数据和在线协作话语的分析,研究发现学习者的提问能力对认知过程存在影响,具体表现为:高提问能力学习者表现出更高水平的智能信息整合能力,能够更多地转述和应用从智能系统中获得的意见;高提问能力学习者呈现出“共识解释发展”的认知建构方式,而低提问能力学习者呈现出“辩证论证”的认知建构方式;高提问能力学习者展现出持续的高认知水平,而低提问能力学习者在协作启动阶段容易陷入无关话题困扰,讨论效率不高。基于此,文章提出了提升学习者人工智能素养与提问能力、细化共识解释与辩证论证教学策略以及开展智能协作学习环境下的纵向研究等建议。