The performance of data restore is one of the key indicators of user experience for backup storage systems.Compared to the traditional offline restore process,online restore reduces downtime during backup restoration,...The performance of data restore is one of the key indicators of user experience for backup storage systems.Compared to the traditional offline restore process,online restore reduces downtime during backup restoration,allowing users to operate on already restored files while other files are still being restored.This approach improves availability during restoration tasks but suffers from a critical limitation:inconsistencies between the access sequence and the restore sequence.In many cases,the file a user needs to access at a given moment may not yet be restored,resulting in significant delays and poor user experience.To this end,we present Histore,which builds on the user’s historical access sequence to schedule the restore sequence,in order to reduce users’access delayed time.Histore includes three restore approaches:(i)the frequency-based approach,which restores files based on historical file access frequencies and prioritizes ensuring the availability of frequently accessed files;(ii)the graph-based approach,which preferentially restores the frequently accessed files as well as their correlated files based on historical access patterns,and(iii)the trie-based approach,which restores particular files based on both users’real-time and historical access patterns to deduce and restore the files to be accessed in the near future.We implement a prototype of Histore and evaluate its performance from multiple perspectives.Trace-driven experiments on two datasets show that Histore significantly reduces users’delay time by 4-700×with only 1.0%-14.5%additional performance overhead.展开更多
Since the initiation of the subseasonal-to-seasonal prediction project by the World Meteorological Organization,the accuracy of model forecasts has improved notably.However,substantial discrepancies have been observed...Since the initiation of the subseasonal-to-seasonal prediction project by the World Meteorological Organization,the accuracy of model forecasts has improved notably.However,substantial discrepancies have been observed among forecast results produced by different ensemble members when applied to South China.To enhance the accuracy of sub-seasonal forecasts in this region,it is essential to develop new methods that can effectively leverage multiple predictive models.This study introduces a weighted ensemble forecasting method based on online learning to improve forecast accuracy.We utilized ensemble forecasts from three models:the Integrated Forecasting System model from the European Centre for Medium-Range Weather Forecasts,the Climate Forecast System Version 2 model from the National Centers for Environmental Prediction,and the Beijing Climate Center-Climate Prediction System version 3 model from the China Meteorological Administration.The ensemble weights are trained using an online learning approach.The results indicate that the forecasts obtained through online learning outperform those of the original dynamical models.Compared to the simple ensemble results of the three models,the weighted ensemble model showed a stronger capability to capture temperature and precipitation patterns in South China.Therefore,this method has the potential to improve the accuracy of sub-seasonal forecasts in this region.展开更多
Topological information is very important for understanding different types of online web services,in particular,for online social networks(OSNs).People leverage such information for various applications,such as socia...Topological information is very important for understanding different types of online web services,in particular,for online social networks(OSNs).People leverage such information for various applications,such as social relationship modeling,community detection,user profiling,and user behavior prediction.However,the leak of such information will also pose severe challenges for user privacy preserving due to its usefulness in characterizing users.Large-scale web crawling-based information probing is a representative way for obtaining topological information of online web services.In this paper,we explore how to defend against topological information probing for online web services,with a particular focus on online decentralized web services such as Mastodon.Different from traditional centralized web services,the federated nature of decentralized web services makes the identification of distributed crawlers even more difficult.We analyze the behavioral differences between legitimate users and crawlers in decentralized web services and highlight two key behavioral attributes that distinguish crawlers from legitimate users:instance interaction preferences and hop count in profile viewing patterns.Based on these insights:we propose a supervised machine learning-based framework for crawler detection,which is able to learn the federation-aware feature representations for users.To validate the framework’s effectiveness,we construct a labeled dataset that integrates real users with real-trace driven simulated crawlers in Mastodon.We use this dataset to train various supervised classifiers for crawler detection.Experimental results demonstrate that our framework can achieve an excellent classification performance.Moreover,it is observed that federation-aware features are effective in improving detection performance.展开更多
Background:Music has proven to be vital in enhancing resilience and promotingwell-being.Previously,the impact of music in sports environments was solely investigated,while this paper applies it to study environments,s...Background:Music has proven to be vital in enhancing resilience and promotingwell-being.Previously,the impact of music in sports environments was solely investigated,while this paper applies it to study environments,standing out as pioneering research.The study consists of a systematic development of a conceptual framework based on theories of Uses and Gratification Expectancy(UGE)and perceived motivation based on music elements.Their components are observed variables influencing students’psychological well-being(as the dependent variable).Resilience is examined as a mediator,influencing the relationships of both observed and dependent variables.The main purpose of this study is to highlight the positive effects of online music consumption on the psychological well-being of students.Methods:Semi-structured qualitative interviews were conducted with eighteen final year creative multimedia undergraduate students belonging to five central region Malaysian universities,especially on their UGE needs,and a similar concept survey instrument with two hundred participants.The interview data were analysed through thematic analysis,while the survey data through descriptive and Partial Least Squares Structural Equation Modeling(PLS-SEM).Results:The results highlight that students gain motivation from online music,which positively affects their psychological well-being(β=0.190,p=0.003,f^(2)=0.037),while resilience significantly affects this relationship(β=0.562,p<0.001,f^(2)=0.461).However,the results also predict a partial relationship between constructs based on UGE with psychological well-being,mediated by resilience,i.e.,AT-UGE(β=0.021,p=0.783,f^(2)=0.000),SIPI-UGE(β=0.228,p=0.004,f^(2)=0.044).Conclusion:The outcome of the study reflected practical,meaningful,and statistically significant results.The majority of the predictors,with the exception of one,i.e.,AT-UGE,displayed a clear positive relation of online music consumption on the Psychological Well-being of students.Future research will explore varying contextual factors impacting online music-related gratifications,motivations,and resilience,along with additional potential mediators and moderators.展开更多
Online Public Opinion Reports consolidate news and social media for timely crisis management by governments and enterprises.While large language models(LLMs)enable automated report generation,this specific domain lack...Online Public Opinion Reports consolidate news and social media for timely crisis management by governments and enterprises.While large language models(LLMs)enable automated report generation,this specific domain lacks formal task definitions and corresponding benchmarks.To bridge this gap,we define the Automated Online Public Opinion Report Generation(OPOR-Gen)task and construct OPOR-Bench,an event-centric dataset with 463 crisis events across 108 countries(comprising 8.8 K news articles and 185 K tweets).To evaluate report quality,we propose OPOR-Eval,a novel agent-based framework that simulates human expert evaluation.Validation experiments show OPOR-Eval achieves a high Spearman’s correlation(ρ=0.70)with human judgments,though challenges in temporal reasoning persist.This work establishes an initial foundation for advancing automated public opinion reporting research.展开更多
An online exhibition of pain tings,by women artists,was held recently.Through the excellent paintings,the artists,also board members of China Female Artists Association,extolled the magnificent views of the motherland...An online exhibition of pain tings,by women artists,was held recently.Through the excellent paintings,the artists,also board members of China Female Artists Association,extolled the magnificent views of the motherland,and they expressed their love and expectations for a better life,as well as their dedication to artistic creation.展开更多
Online examinations have become a dominant assessment mode,increasing concerns over academic integrity.To address the critical challenge of detecting cheating behaviours,this study proposes a hybrid deep learning appr...Online examinations have become a dominant assessment mode,increasing concerns over academic integrity.To address the critical challenge of detecting cheating behaviours,this study proposes a hybrid deep learning approach that combines visual detection and temporal behaviour classification.The methodology utilises object detection models—You Only Look Once(YOLOv12),Faster Region-based Convolutional Neural Network(RCNN),and Single Shot Detector(SSD)MobileNet—integrated with classification models such as Convolutional Neural Networks(CNN),Bidirectional Gated Recurrent Unit(Bi-GRU),and CNN-LSTM(Long Short-Term Memory).Two distinct datasets were used:the Online Exam Proctoring(EOP)dataset from Michigan State University and the School of Computer Science,Duy Tan Unievrsity(SCS-DTU)dataset collected in a controlled classroom setting.A diverse set of cheating behaviours,including book usage,unauthorised interaction,internet access,and mobile phone use,was categorised.Comprehensive experiments evaluated the models based on accuracy,precision,recall,training time,inference speed,and memory usage.We evaluate nine detector-classifier pairings under a unified budget and score them via a calibrated harmonic mean of detection and classification accuracies,enabling deployment-oriented selection under latency and memory constraints.Macro-Precision/Recall/F1 and Receiver Operating Characteristic-Area Under the Curve(ROC-AUC)are reported for the top configurations,revealing consistent advantages of object-centric pipelines for fine-grained cheating cues.The highest overall score is achieved by YOLOv12+CNN(97.15%accuracy),while SSD-MobileNet+CNN provides the best speed-efficiency trade-off for edge devices.This research provides valuable insights into selecting and deploying appropriate deep learning models for maintaining exam integrity under varying resource constraints.展开更多
Improvements in aero-engine performance have made the structures of the aero-engine components increasingly complex.To better adapt to the processing requirements of narrow twisted channels such as an integral shroude...Improvements in aero-engine performance have made the structures of the aero-engine components increasingly complex.To better adapt to the processing requirements of narrow twisted channels such as an integral shrouded blisk,this study proposes an innovative method of electrochemical cutting in which a flexible tube electrode is controlled by online deformation during processing.In this study,the processing principle of electrochemical cutting with a flexible electrode for controlled online deformation(FECC)was revealed for the first time.The online deformation process of flexible electrodes and the machining process of profiles were analysed in depth,and the corresponding theoretical models were established.Conventional electrochemical machining(ECM)is a multi-physical field-coupled process involving electric and flow fields.In FECC,classical mechanics are introduced into the tool cathode,which must be loaded at all times during the machining process.Therefore,in this study,before and after the deformation of the flexible electrode,a corresponding simulation study was conducted to understand the influence of the online deformation of the flexible electrode on the flow and electric fields.The feasibility of flexible electrodes for online deformation and the validity of the theoretical model were verified by deformation measurements and in situ observation experiments.Finally,the method was successfully applied to the machining of nickel-based high-temperature alloys,and different specifications of flexible electrodes were used to complete the machining of the corresponding complex profiles,thereby verifying the feasibility and versatility of the method.The method proposed in this study breaks the tradition of using a non-deformable cathode for ECM and adopts a flexible electrode that can be deformed during the machining process as the tool cathode,which improves machining flexibility and provides a valuable reference to promote the ECM of complex profiles.展开更多
Fuel cell electric vehicles hold great promise for a diverse range of applications in reducing greenhouse gas emissions.In power fuel cell systems,hydrogen fuel serves as an energy vector.To ensure its suitability,it ...Fuel cell electric vehicles hold great promise for a diverse range of applications in reducing greenhouse gas emissions.In power fuel cell systems,hydrogen fuel serves as an energy vector.To ensure its suitability,it is necessary for the quality of hydrogen to adhere to the standards set by ISO 14687:2019,which sets maximum limits for 14 impurities in hydrogen,aiming to prevent any degradation of fuel cell performance.Ammonia(NH_(3))is a prominent pollutant in fuel cells,and accurate measurements of its concentration are crucial for hydrogen fuel cell quantity.In this study,a novel detection platform was developed for determining NH_(3)in real hydrogen samples.The online analysis platform integrates a self-developed online dilution module with a Fourier transform infrared spectrometer(ODM-FTIR).The ODM-FTIR can be operated fully automatically with remote operation.Under the optimum conditions,this method achieved a wide linear range between(50∼1000)nmol/mol.The limit of detection(LOD)was as low as 2 nmol/mol with a relative standard deviation(RSD,n=7)of 3.6%at a content of 50 nmol/mol.To ensure that the quality of the hydrogen products meets the requirement of proton exchange membrane fuel cell vehicles(PEMFCV),the developed ODM-FTIR system was applied to monitor the NH_(3)content in Chengdu Hydrogen Energy Co.,Ltd.for 21 days during Chengdu 2021 FISU World University Games.The proposed method retains several unique advantages,including a low detection limit,excellent repeatability,high accuracy,high speed,good stability,and calibration flexibility.It is an effective analytical method for accurately quantifying NH_(3)in hydrogen,especially suitable for online analysis.It also provides a new idea for the analysis of other impurity components in hydrogen.展开更多
To address the issue of coordinated control of multiple hydrogen and battery storage units to suppress the grid-injected power deviation of wind farms,an online optimization strategy for Battery-hydrogen hybrid energy...To address the issue of coordinated control of multiple hydrogen and battery storage units to suppress the grid-injected power deviation of wind farms,an online optimization strategy for Battery-hydrogen hybrid energy storage systems based on measurement feedback is proposed.First,considering the high charge/discharge losses of hydrogen storage and the low energy density of battery storage,an operational optimization objective is established to enable adaptive energy adjustment in the Battery-hydrogen hybrid energy storage system.Next,an online optimization model minimizing the operational cost of the hybrid system is constructed to suppress grid-injected power deviations with satisfying the operational constraints of hydrogen storage and batteries.Finally,utilizing the online measurement of the energy states of hydrogen storage and batteries,an online optimization strategy based on measurement feedback is designed.Case study results show:before and after smoothing the fluctuations in wind power,the time when the power exceeded the upper and lower limits of the grid-injected power accounted for 24.1%and 1.45%of the total time,respectively,the proposed strategy can effectively keep the grid-injected power deviations of wind farms within the allowable range.Hydrogen storage and batteries respectively undertake long-term and short-term charge/discharge tasks,effectively reducing charge/discharge losses of the Battery-hydrogen hybrid energy storage systems and improving its operational efficiency.展开更多
In order to address the current inability of screen printing to monitor printing pressure online,an online printing pressure monitoring system applied to screen printing machines was designed in this study.In this stu...In order to address the current inability of screen printing to monitor printing pressure online,an online printing pressure monitoring system applied to screen printing machines was designed in this study.In this study,the consistency of printed electrodes was measured by using a confocal microscope and the pressure distribution detected by online pressure monitoring system was compared to investigate the relationship.The results demonstrated the relationship between printing pressure and the consistency of printed electrodes.As printing pressure increases,the ink layer at the corresponding position becomes thicker and that higher printing pressure enhances the consistency of the printed electrodes.The experiment confirms the feasibility of the online pressure monitoring system,which aids in predicting and controlling the consistency of printed electrodes,thereby improving their performance.展开更多
Accurately determining when and what to remanufacture is essential for maximizing the lifecycle value of industrial equipment.However,existing approaches face three significant limitations:(1)reliance on predefined ma...Accurately determining when and what to remanufacture is essential for maximizing the lifecycle value of industrial equipment.However,existing approaches face three significant limitations:(1)reliance on predefined mathematical models that often fail to capture equipment-specific degradation,(2)offline optimization methods that assume access to future data,and(3)the absence of component-level guidance.To address these challenges,we propose a data-driven framework for component-level decision-making.The framework leverages streaming sensor data to predict the remaining useful life(RUL)without relying on mathematical models,employs an online optimization algorithm suitable for practical settings,and,through remanufacturing simulations,provides guidance on which components should be replaced.In a case study on gas-insulated switchgear,the proposed framework achieved RUL prediction performance comparable to an oracle model in an online setting without relying on predefined mathematical models.Furthermore,by employing online optimization,it determined a remanufacturing timing close to the global optimum using only past and current data.In addition,unlike previous studies,the framework enables component-level decision-making,allowing for more detailed and actionable remanufacturing guidance in practical applications.展开更多
In this modern era, platforms for digital/social media and video games are growing daily. People are becoming dependent on them from all ages and with many positive aspects, but there are drawbacks as well, one of whi...In this modern era, platforms for digital/social media and video games are growing daily. People are becoming dependent on them from all ages and with many positive aspects, but there are drawbacks as well, one of which is cyberbullying. Cyberbullying is a form of bullying that uses technological platforms to bully others. It has effects on victims mentally, emotionally, and physically, which include low self-esteem, acting violently, despair, increased stress/anxiety, depression, self-harming/suicide, etc. Findings from this research study justify that it affects young people more, impacting their emotional development and overall safety. Real-time cyberbullying detection identifies and protects the target from further abuse and its effects. This study aids in determining the seriousness of the issue and the vulnerabilities that individuals can take advantage of to bully others. Additionally, it will help to understand how various features of cyberbullying detection function assist in developing a strong and trustworthy system and making a healthy online community. Natural Language Processing (NLP) models assess the textual content and analyze hashtags and comments. Similarly, image context is analyzed using Optical Character Recognition (OCR), which converts images into a machine-readable format for further examination. There are also Deep Neural Network models, such as Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), and Bidirectional LSTM (BLSTM). CNN is utilized for text/picture classification, LSTM is used for long-term dependency learning, and BLSTM expands the network’s input by encoding data in both forward and backward directions. Classifiers like Support Vector Machine (SVM) and Naïve Bayes help detect cyberbullying. A working cyberbullying detection system can detect cyberbullying on multiple platforms. A deeper understanding of each machine learning algorithm allows one to build a model that improves upon their predecessors. With models being developed for different attributes providing results with high accuracy, the cyberbullying detection system contributes by leading us to a healthier online community.展开更多
With the rapid development of the internet,the dissemination of public opinion in online social networks has become increasingly complex.Existing dissemination models rarely consider group phenomena and the simultaneo...With the rapid development of the internet,the dissemination of public opinion in online social networks has become increasingly complex.Existing dissemination models rarely consider group phenomena and the simultaneous spread of competing public opinion information in online social networks.This paper introduces the UHNPR information dissemination model to study the dynamic spread and interaction of positive and negative public opinion information in hypernetworks.To improve the accuracy of modeling of information dissemination,we revise the traditional assumptions of constant propagation and decay rates by redefining these rates based on factors that influence the spread of public opinion information.Subsequently,we validate the effectiveness of the UHNPR model using numerical simulations and analyze the impact of factors such as authority effect,user intimacy,information content and information timeliness on the spread of public opinion,providing corresponding suggestions for public opinion control.Our research results demonstrate that this model outperforms the SIR,SEIR and SEIDR models in describing public opinion propagation in real social networks.Compared with complex networks,information spreads faster and more extensively in hypernetworks.展开更多
The operation furnace profile for the high heat load zone was one of the important factors affecting the stable and high-quality production of the blast furnace,but it was difficult to monitor directly.To address this...The operation furnace profile for the high heat load zone was one of the important factors affecting the stable and high-quality production of the blast furnace,but it was difficult to monitor directly.To address this issue,an online calculation model for the operation furnace profile was proposed based on a dual-driven approach combining data and mechanisms,by integrating mechanism experiment,numerical simulation,and machine learning.The experimentally determined slag layer hanging temperature was 1130℃,and the thermal conductivity ranged from 1.32 to 1.96 m^(2)℃^(-1).Based on the 3D slag-hanging numerical simulation model,a database was constructed,containing 2294 sets of mechanism cases for the slag layer.The fusion of data modeling,heat transfer theory,and expert experience enabled the online calculation of key input variables for the operation furnace profile,particularly the quantification of the“black-box”variable of gas temperature.Simulated data were used as inputs,and light gradient boosting machine was applied to construct the online calculation model for the operation furnace profile.This model facilitated the online calculation of the slag layer thickness and other key indices.The coefficient of determination of the model exceeded 0.98,indicating high accuracy.A slag layer state judgment model was constructed,categorizing states as shedding,too thin,normal,and too thick.Real-time data were applied,and the average slag thickness in the high heat load area of the test data ranged from 40 to 80 mm,which was consistent with field experience.The absolute value of the Pearson correlation coefficient between slag layer thickness,thermocouple temperature,and heat load data was above 0.85,indicating that the calculated results closely aligned with the actual trends.A 3D visual online monitoring system for the operation furnace profile was created,and it has been successfully implemented at the blast furnace site.展开更多
On November 21,2024,a book collection ceremony that connects the east and west was held under the red brick dome of the British Library in London.Ten literary works,originally published online by Chinese authors,inclu...On November 21,2024,a book collection ceremony that connects the east and west was held under the red brick dome of the British Library in London.Ten literary works,originally published online by Chinese authors,including“The Lord of the Mysteries”,“The Ring of Destiny”,“We Live in Nanjing”and“Zanxing”were officially added to the collection of the library.From the“online literature hits”that attracted countless people to flood the Internet in the depth of the night to the“classic works”collected by a famous foreign library,Chinese online literature works have become another testimony to the cultural exchanges between China and foreign countries.Behind this“coming-from-behind”phenomenon is the rapid development of Chinese online literature over the past 20 years-from the barbaric growth of local grassroots culture to the“new quality productive force”of global cultural trade.Its evolutionary trajectory is full of romanticism and also contains deep propositions of industrial transformation.展开更多
With the rapid development of online education,the impact of interface design on learning experience has become increasingly prominent.Reasonable color matching can effectively improve learning efficiency,enhance user...With the rapid development of online education,the impact of interface design on learning experience has become increasingly prominent.Reasonable color matching can effectively improve learning efficiency,enhance user engagement,and improve visual experience.This paper analyzes the application of color matching in interface design,discusses the principle of color matching in online course interfaces,and puts forward some design strategies.It provides a practical reference for the interface design of an online education platform.展开更多
Dear Editor,This letter presents a novel approach to the data-driven control of unknown nonlinear systems.By leveraging online sparse identification based on the Koopman operator,a high-dimensional linear system model...Dear Editor,This letter presents a novel approach to the data-driven control of unknown nonlinear systems.By leveraging online sparse identification based on the Koopman operator,a high-dimensional linear system model approximating the actual system is obtained online.The upper bound of the discrepancy between the identified model and the actual system is estimated using real-time prediction error,which is then utilized in the design of a tube-based robust model predictive controller.The effectiveness of the proposed approach is validated by numerical simulation.展开更多
Background/Need for innovation: Undergraduate students in Otolaryngology are on the lookout for easy modes of learning which can help them understand concepts better as well as score more in examinations. A need was h...Background/Need for innovation: Undergraduate students in Otolaryngology are on the lookout for easy modes of learning which can help them understand concepts better as well as score more in examinations. A need was hence felt to introduce a new learning resource to supplement traditional teaching-learning methods. Methods: Digital, case based self–study modules were prepared using all open source technology and validated by experts in the specialty. The modules were uploaded on a website specifically created for the purpose. They were pilot tested on twenty consenting third year undergraduate (MBBS) students using a crossover design. Post test comprising of multiple choice questions was administered to the students after a period of two weeks. Feedback was obtained from faculty and students. Results: Test scores were found to be significantly higher amongst students who used the learning modules as a supplement to regular bedside teaching (p < 0.001;Wilcoxon signed rank test). Majority of students agreed that the modules helped them gain confidence during internal assessment examinations and would help revision. Conclusions: Online, case based, self-study modules helped students to perform better when used as a supplement to traditional teaching methods. Students agreed that it enabled easy understanding of subject and helped them gain confidence.展开更多
Online education has become a critical mode of instruction in Chinese universities,particularly during and after the COVID-19 pandemic.This study examines information management in online education through a triadic f...Online education has become a critical mode of instruction in Chinese universities,particularly during and after the COVID-19 pandemic.This study examines information management in online education through a triadic framework encompassing classroom information management,teaching effectiveness management,and teaching information management.Drawing on in-depth interviews with administrators,teachers,and students,the findings reveal three primary challenges in online education:the absence of embodied information,the uncontrollable nature of online platforms,and nonverbal overload caused by Zoom fatigue.Teachers face difficulties maintaining presence and interaction due to limited feedback from students and risks associated with class recordings.Students experience increased psychological and physical fatigue due to the overlap of learning and living spaces.Recommendations to address these issues include enhancing teacher-student interaction to foster a sense of presence,improving the transparency of course information to align expectations,and adopting user-friendly teaching platforms with privacy safeguards.These insights aim to improve the security,effectiveness,and experience of online education in higher education institutions.展开更多
基金supported in part by National Key R&D Program of China(2022YFB4501200),National Natural Science Foundation of China(62332018)Science and Technology Program(2024NSFTD0031,2024YFHZ0339 and 2025ZNSFSC0497).
文摘The performance of data restore is one of the key indicators of user experience for backup storage systems.Compared to the traditional offline restore process,online restore reduces downtime during backup restoration,allowing users to operate on already restored files while other files are still being restored.This approach improves availability during restoration tasks but suffers from a critical limitation:inconsistencies between the access sequence and the restore sequence.In many cases,the file a user needs to access at a given moment may not yet be restored,resulting in significant delays and poor user experience.To this end,we present Histore,which builds on the user’s historical access sequence to schedule the restore sequence,in order to reduce users’access delayed time.Histore includes three restore approaches:(i)the frequency-based approach,which restores files based on historical file access frequencies and prioritizes ensuring the availability of frequently accessed files;(ii)the graph-based approach,which preferentially restores the frequently accessed files as well as their correlated files based on historical access patterns,and(iii)the trie-based approach,which restores particular files based on both users’real-time and historical access patterns to deduce and restore the files to be accessed in the near future.We implement a prototype of Histore and evaluate its performance from multiple perspectives.Trace-driven experiments on two datasets show that Histore significantly reduces users’delay time by 4-700×with only 1.0%-14.5%additional performance overhead.
基金Science and Technology Development Program of the“Taihu Light”(K20231023)CMA Numerical Weather Prediction R&D Project(TCYF2024QH007)+1 种基金“Qing Lan”Project of Jiangsu Province for C.H.LUWuxi University Research Start-up Fund for Introduced Talents(2023r037)。
文摘Since the initiation of the subseasonal-to-seasonal prediction project by the World Meteorological Organization,the accuracy of model forecasts has improved notably.However,substantial discrepancies have been observed among forecast results produced by different ensemble members when applied to South China.To enhance the accuracy of sub-seasonal forecasts in this region,it is essential to develop new methods that can effectively leverage multiple predictive models.This study introduces a weighted ensemble forecasting method based on online learning to improve forecast accuracy.We utilized ensemble forecasts from three models:the Integrated Forecasting System model from the European Centre for Medium-Range Weather Forecasts,the Climate Forecast System Version 2 model from the National Centers for Environmental Prediction,and the Beijing Climate Center-Climate Prediction System version 3 model from the China Meteorological Administration.The ensemble weights are trained using an online learning approach.The results indicate that the forecasts obtained through online learning outperform those of the original dynamical models.Compared to the simple ensemble results of the three models,the weighted ensemble model showed a stronger capability to capture temperature and precipitation patterns in South China.Therefore,this method has the potential to improve the accuracy of sub-seasonal forecasts in this region.
基金funded by the National Key R&D Program of China under Grant(No.2022YFB3102901)National Natural Science Foundation of China(No.62072115,No.62102094)Shanghai Science and Technology Innovation Action Plan Project(No.22510713600).
文摘Topological information is very important for understanding different types of online web services,in particular,for online social networks(OSNs).People leverage such information for various applications,such as social relationship modeling,community detection,user profiling,and user behavior prediction.However,the leak of such information will also pose severe challenges for user privacy preserving due to its usefulness in characterizing users.Large-scale web crawling-based information probing is a representative way for obtaining topological information of online web services.In this paper,we explore how to defend against topological information probing for online web services,with a particular focus on online decentralized web services such as Mastodon.Different from traditional centralized web services,the federated nature of decentralized web services makes the identification of distributed crawlers even more difficult.We analyze the behavioral differences between legitimate users and crawlers in decentralized web services and highlight two key behavioral attributes that distinguish crawlers from legitimate users:instance interaction preferences and hop count in profile viewing patterns.Based on these insights:we propose a supervised machine learning-based framework for crawler detection,which is able to learn the federation-aware feature representations for users.To validate the framework’s effectiveness,we construct a labeled dataset that integrates real users with real-trace driven simulated crawlers in Mastodon.We use this dataset to train various supervised classifiers for crawler detection.Experimental results demonstrate that our framework can achieve an excellent classification performance.Moreover,it is observed that federation-aware features are effective in improving detection performance.
基金funded by Malaysian Ministry of Higher Education(MOHE)under the Fundamental Research Grant Scheme(FRGS/1/2023/SSI07/MMU/02/3)which is awarded to the Multimedia University.The project is led by the second author.
文摘Background:Music has proven to be vital in enhancing resilience and promotingwell-being.Previously,the impact of music in sports environments was solely investigated,while this paper applies it to study environments,standing out as pioneering research.The study consists of a systematic development of a conceptual framework based on theories of Uses and Gratification Expectancy(UGE)and perceived motivation based on music elements.Their components are observed variables influencing students’psychological well-being(as the dependent variable).Resilience is examined as a mediator,influencing the relationships of both observed and dependent variables.The main purpose of this study is to highlight the positive effects of online music consumption on the psychological well-being of students.Methods:Semi-structured qualitative interviews were conducted with eighteen final year creative multimedia undergraduate students belonging to five central region Malaysian universities,especially on their UGE needs,and a similar concept survey instrument with two hundred participants.The interview data were analysed through thematic analysis,while the survey data through descriptive and Partial Least Squares Structural Equation Modeling(PLS-SEM).Results:The results highlight that students gain motivation from online music,which positively affects their psychological well-being(β=0.190,p=0.003,f^(2)=0.037),while resilience significantly affects this relationship(β=0.562,p<0.001,f^(2)=0.461).However,the results also predict a partial relationship between constructs based on UGE with psychological well-being,mediated by resilience,i.e.,AT-UGE(β=0.021,p=0.783,f^(2)=0.000),SIPI-UGE(β=0.228,p=0.004,f^(2)=0.044).Conclusion:The outcome of the study reflected practical,meaningful,and statistically significant results.The majority of the predictors,with the exception of one,i.e.,AT-UGE,displayed a clear positive relation of online music consumption on the Psychological Well-being of students.Future research will explore varying contextual factors impacting online music-related gratifications,motivations,and resilience,along with additional potential mediators and moderators.
基金supported by the Fundamental Research Funds for the Central Universities(No.CUC25SG013)the Foundation of Key Laboratory of Education Informatization for Nationalities(Yunnan Normal University),Ministry of Education(No.EIN2024C006).
文摘Online Public Opinion Reports consolidate news and social media for timely crisis management by governments and enterprises.While large language models(LLMs)enable automated report generation,this specific domain lacks formal task definitions and corresponding benchmarks.To bridge this gap,we define the Automated Online Public Opinion Report Generation(OPOR-Gen)task and construct OPOR-Bench,an event-centric dataset with 463 crisis events across 108 countries(comprising 8.8 K news articles and 185 K tweets).To evaluate report quality,we propose OPOR-Eval,a novel agent-based framework that simulates human expert evaluation.Validation experiments show OPOR-Eval achieves a high Spearman’s correlation(ρ=0.70)with human judgments,though challenges in temporal reasoning persist.This work establishes an initial foundation for advancing automated public opinion reporting research.
文摘An online exhibition of pain tings,by women artists,was held recently.Through the excellent paintings,the artists,also board members of China Female Artists Association,extolled the magnificent views of the motherland,and they expressed their love and expectations for a better life,as well as their dedication to artistic creation.
文摘Online examinations have become a dominant assessment mode,increasing concerns over academic integrity.To address the critical challenge of detecting cheating behaviours,this study proposes a hybrid deep learning approach that combines visual detection and temporal behaviour classification.The methodology utilises object detection models—You Only Look Once(YOLOv12),Faster Region-based Convolutional Neural Network(RCNN),and Single Shot Detector(SSD)MobileNet—integrated with classification models such as Convolutional Neural Networks(CNN),Bidirectional Gated Recurrent Unit(Bi-GRU),and CNN-LSTM(Long Short-Term Memory).Two distinct datasets were used:the Online Exam Proctoring(EOP)dataset from Michigan State University and the School of Computer Science,Duy Tan Unievrsity(SCS-DTU)dataset collected in a controlled classroom setting.A diverse set of cheating behaviours,including book usage,unauthorised interaction,internet access,and mobile phone use,was categorised.Comprehensive experiments evaluated the models based on accuracy,precision,recall,training time,inference speed,and memory usage.We evaluate nine detector-classifier pairings under a unified budget and score them via a calibrated harmonic mean of detection and classification accuracies,enabling deployment-oriented selection under latency and memory constraints.Macro-Precision/Recall/F1 and Receiver Operating Characteristic-Area Under the Curve(ROC-AUC)are reported for the top configurations,revealing consistent advantages of object-centric pipelines for fine-grained cheating cues.The highest overall score is achieved by YOLOv12+CNN(97.15%accuracy),while SSD-MobileNet+CNN provides the best speed-efficiency trade-off for edge devices.This research provides valuable insights into selecting and deploying appropriate deep learning models for maintaining exam integrity under varying resource constraints.
基金supported by the National Natural Science Foundation of China(52375443)the Innovative Research Group Project of the National Natural Science Foundation of China(51921003).
文摘Improvements in aero-engine performance have made the structures of the aero-engine components increasingly complex.To better adapt to the processing requirements of narrow twisted channels such as an integral shrouded blisk,this study proposes an innovative method of electrochemical cutting in which a flexible tube electrode is controlled by online deformation during processing.In this study,the processing principle of electrochemical cutting with a flexible electrode for controlled online deformation(FECC)was revealed for the first time.The online deformation process of flexible electrodes and the machining process of profiles were analysed in depth,and the corresponding theoretical models were established.Conventional electrochemical machining(ECM)is a multi-physical field-coupled process involving electric and flow fields.In FECC,classical mechanics are introduced into the tool cathode,which must be loaded at all times during the machining process.Therefore,in this study,before and after the deformation of the flexible electrode,a corresponding simulation study was conducted to understand the influence of the online deformation of the flexible electrode on the flow and electric fields.The feasibility of flexible electrodes for online deformation and the validity of the theoretical model were verified by deformation measurements and in situ observation experiments.Finally,the method was successfully applied to the machining of nickel-based high-temperature alloys,and different specifications of flexible electrodes were used to complete the machining of the corresponding complex profiles,thereby verifying the feasibility and versatility of the method.The method proposed in this study breaks the tradition of using a non-deformable cathode for ECM and adopts a flexible electrode that can be deformed during the machining process as the tool cathode,which improves machining flexibility and provides a valuable reference to promote the ECM of complex profiles.
基金financial support by Sichuan Science and Technology,China(No.2023YFG0070).
文摘Fuel cell electric vehicles hold great promise for a diverse range of applications in reducing greenhouse gas emissions.In power fuel cell systems,hydrogen fuel serves as an energy vector.To ensure its suitability,it is necessary for the quality of hydrogen to adhere to the standards set by ISO 14687:2019,which sets maximum limits for 14 impurities in hydrogen,aiming to prevent any degradation of fuel cell performance.Ammonia(NH_(3))is a prominent pollutant in fuel cells,and accurate measurements of its concentration are crucial for hydrogen fuel cell quantity.In this study,a novel detection platform was developed for determining NH_(3)in real hydrogen samples.The online analysis platform integrates a self-developed online dilution module with a Fourier transform infrared spectrometer(ODM-FTIR).The ODM-FTIR can be operated fully automatically with remote operation.Under the optimum conditions,this method achieved a wide linear range between(50∼1000)nmol/mol.The limit of detection(LOD)was as low as 2 nmol/mol with a relative standard deviation(RSD,n=7)of 3.6%at a content of 50 nmol/mol.To ensure that the quality of the hydrogen products meets the requirement of proton exchange membrane fuel cell vehicles(PEMFCV),the developed ODM-FTIR system was applied to monitor the NH_(3)content in Chengdu Hydrogen Energy Co.,Ltd.for 21 days during Chengdu 2021 FISU World University Games.The proposed method retains several unique advantages,including a low detection limit,excellent repeatability,high accuracy,high speed,good stability,and calibration flexibility.It is an effective analytical method for accurately quantifying NH_(3)in hydrogen,especially suitable for online analysis.It also provides a new idea for the analysis of other impurity components in hydrogen.
基金Supported by State Grid Zhejiang Electric Power Co.,Ltd.Science and Technology Project Funding(No.B311DS230005).
文摘To address the issue of coordinated control of multiple hydrogen and battery storage units to suppress the grid-injected power deviation of wind farms,an online optimization strategy for Battery-hydrogen hybrid energy storage systems based on measurement feedback is proposed.First,considering the high charge/discharge losses of hydrogen storage and the low energy density of battery storage,an operational optimization objective is established to enable adaptive energy adjustment in the Battery-hydrogen hybrid energy storage system.Next,an online optimization model minimizing the operational cost of the hybrid system is constructed to suppress grid-injected power deviations with satisfying the operational constraints of hydrogen storage and batteries.Finally,utilizing the online measurement of the energy states of hydrogen storage and batteries,an online optimization strategy based on measurement feedback is designed.Case study results show:before and after smoothing the fluctuations in wind power,the time when the power exceeded the upper and lower limits of the grid-injected power accounted for 24.1%and 1.45%of the total time,respectively,the proposed strategy can effectively keep the grid-injected power deviations of wind farms within the allowable range.Hydrogen storage and batteries respectively undertake long-term and short-term charge/discharge tasks,effectively reducing charge/discharge losses of the Battery-hydrogen hybrid energy storage systems and improving its operational efficiency.
文摘In order to address the current inability of screen printing to monitor printing pressure online,an online printing pressure monitoring system applied to screen printing machines was designed in this study.In this study,the consistency of printed electrodes was measured by using a confocal microscope and the pressure distribution detected by online pressure monitoring system was compared to investigate the relationship.The results demonstrated the relationship between printing pressure and the consistency of printed electrodes.As printing pressure increases,the ink layer at the corresponding position becomes thicker and that higher printing pressure enhances the consistency of the printed electrodes.The experiment confirms the feasibility of the online pressure monitoring system,which aids in predicting and controlling the consistency of printed electrodes,thereby improving their performance.
基金supported by the Human Resources Development of the Korea Institute of Energy Technology Evaluation and Planning(KETEP)grant funded by the Korea government Ministry of Knowledge Economy(No.RS-2023-00244330)the National Research Foundation of Korea(NRF)grant funded by the Korea government(MSIT)(No.NRF RS-2023-00219052RS-2024-00352587)。
文摘Accurately determining when and what to remanufacture is essential for maximizing the lifecycle value of industrial equipment.However,existing approaches face three significant limitations:(1)reliance on predefined mathematical models that often fail to capture equipment-specific degradation,(2)offline optimization methods that assume access to future data,and(3)the absence of component-level guidance.To address these challenges,we propose a data-driven framework for component-level decision-making.The framework leverages streaming sensor data to predict the remaining useful life(RUL)without relying on mathematical models,employs an online optimization algorithm suitable for practical settings,and,through remanufacturing simulations,provides guidance on which components should be replaced.In a case study on gas-insulated switchgear,the proposed framework achieved RUL prediction performance comparable to an oracle model in an online setting without relying on predefined mathematical models.Furthermore,by employing online optimization,it determined a remanufacturing timing close to the global optimum using only past and current data.In addition,unlike previous studies,the framework enables component-level decision-making,allowing for more detailed and actionable remanufacturing guidance in practical applications.
文摘In this modern era, platforms for digital/social media and video games are growing daily. People are becoming dependent on them from all ages and with many positive aspects, but there are drawbacks as well, one of which is cyberbullying. Cyberbullying is a form of bullying that uses technological platforms to bully others. It has effects on victims mentally, emotionally, and physically, which include low self-esteem, acting violently, despair, increased stress/anxiety, depression, self-harming/suicide, etc. Findings from this research study justify that it affects young people more, impacting their emotional development and overall safety. Real-time cyberbullying detection identifies and protects the target from further abuse and its effects. This study aids in determining the seriousness of the issue and the vulnerabilities that individuals can take advantage of to bully others. Additionally, it will help to understand how various features of cyberbullying detection function assist in developing a strong and trustworthy system and making a healthy online community. Natural Language Processing (NLP) models assess the textual content and analyze hashtags and comments. Similarly, image context is analyzed using Optical Character Recognition (OCR), which converts images into a machine-readable format for further examination. There are also Deep Neural Network models, such as Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), and Bidirectional LSTM (BLSTM). CNN is utilized for text/picture classification, LSTM is used for long-term dependency learning, and BLSTM expands the network’s input by encoding data in both forward and backward directions. Classifiers like Support Vector Machine (SVM) and Naïve Bayes help detect cyberbullying. A working cyberbullying detection system can detect cyberbullying on multiple platforms. A deeper understanding of each machine learning algorithm allows one to build a model that improves upon their predecessors. With models being developed for different attributes providing results with high accuracy, the cyberbullying detection system contributes by leading us to a healthier online community.
基金supported by Yunnan High-tech Industry Development Project(Grant No.201606)Yunnan Provincial Major Science and Technology Special Plan Projects(Grant Nos.202103AA080015 and 202002AD080001-5)+1 种基金Yunnan Basic Research Project(Grant No.202001AS070014)Talents and Platform Program of Science and Technology of Yunnan(Grant No.202105AC160018)。
文摘With the rapid development of the internet,the dissemination of public opinion in online social networks has become increasingly complex.Existing dissemination models rarely consider group phenomena and the simultaneous spread of competing public opinion information in online social networks.This paper introduces the UHNPR information dissemination model to study the dynamic spread and interaction of positive and negative public opinion information in hypernetworks.To improve the accuracy of modeling of information dissemination,we revise the traditional assumptions of constant propagation and decay rates by redefining these rates based on factors that influence the spread of public opinion information.Subsequently,we validate the effectiveness of the UHNPR model using numerical simulations and analyze the impact of factors such as authority effect,user intimacy,information content and information timeliness on the spread of public opinion,providing corresponding suggestions for public opinion control.Our research results demonstrate that this model outperforms the SIR,SEIR and SEIDR models in describing public opinion propagation in real social networks.Compared with complex networks,information spreads faster and more extensively in hypernetworks.
基金financially supported by the National Natural Science Foundation of China(Grant Nos.52404343 and 52274326)the Fundamental Research Funds for the Central Universities(Grant Nos.N2425031 and N25BJD007)+1 种基金the China Postdoctoral Science Foundation(Grant No.2024M760370)the Liaoning Province Science and Technology Plan Joint Program(Key Research and Development Program Project)(Grant No.2023JH2/101800058).
文摘The operation furnace profile for the high heat load zone was one of the important factors affecting the stable and high-quality production of the blast furnace,but it was difficult to monitor directly.To address this issue,an online calculation model for the operation furnace profile was proposed based on a dual-driven approach combining data and mechanisms,by integrating mechanism experiment,numerical simulation,and machine learning.The experimentally determined slag layer hanging temperature was 1130℃,and the thermal conductivity ranged from 1.32 to 1.96 m^(2)℃^(-1).Based on the 3D slag-hanging numerical simulation model,a database was constructed,containing 2294 sets of mechanism cases for the slag layer.The fusion of data modeling,heat transfer theory,and expert experience enabled the online calculation of key input variables for the operation furnace profile,particularly the quantification of the“black-box”variable of gas temperature.Simulated data were used as inputs,and light gradient boosting machine was applied to construct the online calculation model for the operation furnace profile.This model facilitated the online calculation of the slag layer thickness and other key indices.The coefficient of determination of the model exceeded 0.98,indicating high accuracy.A slag layer state judgment model was constructed,categorizing states as shedding,too thin,normal,and too thick.Real-time data were applied,and the average slag thickness in the high heat load area of the test data ranged from 40 to 80 mm,which was consistent with field experience.The absolute value of the Pearson correlation coefficient between slag layer thickness,thermocouple temperature,and heat load data was above 0.85,indicating that the calculated results closely aligned with the actual trends.A 3D visual online monitoring system for the operation furnace profile was created,and it has been successfully implemented at the blast furnace site.
文摘On November 21,2024,a book collection ceremony that connects the east and west was held under the red brick dome of the British Library in London.Ten literary works,originally published online by Chinese authors,including“The Lord of the Mysteries”,“The Ring of Destiny”,“We Live in Nanjing”and“Zanxing”were officially added to the collection of the library.From the“online literature hits”that attracted countless people to flood the Internet in the depth of the night to the“classic works”collected by a famous foreign library,Chinese online literature works have become another testimony to the cultural exchanges between China and foreign countries.Behind this“coming-from-behind”phenomenon is the rapid development of Chinese online literature over the past 20 years-from the barbaric growth of local grassroots culture to the“new quality productive force”of global cultural trade.Its evolutionary trajectory is full of romanticism and also contains deep propositions of industrial transformation.
文摘With the rapid development of online education,the impact of interface design on learning experience has become increasingly prominent.Reasonable color matching can effectively improve learning efficiency,enhance user engagement,and improve visual experience.This paper analyzes the application of color matching in interface design,discusses the principle of color matching in online course interfaces,and puts forward some design strategies.It provides a practical reference for the interface design of an online education platform.
基金supported by the National Natural Science Foundation of China(62473020).
文摘Dear Editor,This letter presents a novel approach to the data-driven control of unknown nonlinear systems.By leveraging online sparse identification based on the Koopman operator,a high-dimensional linear system model approximating the actual system is obtained online.The upper bound of the discrepancy between the identified model and the actual system is estimated using real-time prediction error,which is then utilized in the design of a tube-based robust model predictive controller.The effectiveness of the proposed approach is validated by numerical simulation.
文摘Background/Need for innovation: Undergraduate students in Otolaryngology are on the lookout for easy modes of learning which can help them understand concepts better as well as score more in examinations. A need was hence felt to introduce a new learning resource to supplement traditional teaching-learning methods. Methods: Digital, case based self–study modules were prepared using all open source technology and validated by experts in the specialty. The modules were uploaded on a website specifically created for the purpose. They were pilot tested on twenty consenting third year undergraduate (MBBS) students using a crossover design. Post test comprising of multiple choice questions was administered to the students after a period of two weeks. Feedback was obtained from faculty and students. Results: Test scores were found to be significantly higher amongst students who used the learning modules as a supplement to regular bedside teaching (p < 0.001;Wilcoxon signed rank test). Majority of students agreed that the modules helped them gain confidence during internal assessment examinations and would help revision. Conclusions: Online, case based, self-study modules helped students to perform better when used as a supplement to traditional teaching methods. Students agreed that it enabled easy understanding of subject and helped them gain confidence.
文摘Online education has become a critical mode of instruction in Chinese universities,particularly during and after the COVID-19 pandemic.This study examines information management in online education through a triadic framework encompassing classroom information management,teaching effectiveness management,and teaching information management.Drawing on in-depth interviews with administrators,teachers,and students,the findings reveal three primary challenges in online education:the absence of embodied information,the uncontrollable nature of online platforms,and nonverbal overload caused by Zoom fatigue.Teachers face difficulties maintaining presence and interaction due to limited feedback from students and risks associated with class recordings.Students experience increased psychological and physical fatigue due to the overlap of learning and living spaces.Recommendations to address these issues include enhancing teacher-student interaction to foster a sense of presence,improving the transparency of course information to align expectations,and adopting user-friendly teaching platforms with privacy safeguards.These insights aim to improve the security,effectiveness,and experience of online education in higher education institutions.
