教学课程资源库 更多>>

关键词： Computer science   Engineering   Materials science  

摘要： Carbon-capture technology often founders at the point when basic research is translated into practical applications. A computational modelling platform called PrISMa solves this problem by considering the needs of all stakeholders.

摘要： This study evaluates the extent of leisure time loss experienced by middle-aged women in South Korea who care for their aging parents and quantifies the monetary value of this lost leisure time. By employing an open-ended direct question, we derived respondents' willingness to pay and willingness to accept compensation for leisure time lost due to caregiving duties. Our findings reveal that middle-aged women engaged in regular caregiving are willing to pay an average of 9313 KRW (US$7.74) per hour for leisure time and would require 13,539 KRW (US$11.40) per hour as compensation for their lost leisure. The annual total monetary value of leisure time loss for consistent caregivers was estimated to range between 7403.4 billion KRW (US$6232 million) and 10,127.5 billion KRW (US$8525 million). The study highlights the significant challenges faced by caregivers, particularly middle-aged women, emphasizing the need for social awareness and discussions on gender role distribution to promote gender equality. It underscores the importance of developing effective support policies, including financial compensation and mental health resources, to alleviate caregivers' dual burden of work and home responsibilities. Understanding the economic impact of leisure time loss can help address long-term socioeconomic issues associated with an aging society, crucial for improving the efficiency and sustainability of care services for the older generation. The economic value estimated in this study will provide essential information for evaluating the cost-effectiveness of public services, supporting the economic feasibility of policies and programs, especially because respite services involve significant government expenditure.

关键词： Short-term trajectories   Expectancy -value -cost   Chemistry   Computer science   University students   UTILITY-VALUE INTERVENTION   MOTIVATION TRAJECTORIES   UNDERGRADUATE SCIENCE   IDENTITY DEVELOPMENT   ACHIEVEMENT   STUDENTS   CLASSROOM   BELIEFS   MATH   EXPECTATIONS  

摘要： University students studying science, technology, engineering, and mathematics (STEM) tend to lose motivation, on average, over time. However, detailed understanding of when and how students are most likely to lose motivation, as well as whether motivational trajectories differ across contexts, is lacking. Accordingly, we investigated trajectories of expectancy, task values, and perceived costs in four introductory STEM courses (N = 2,104). Latent change score models revealed nuanced patterns of change, indicating average motivational declines in expectancy and task values and average increases in cost perceptions observed in prior research tend to occur primarily at the beginning of the semester. Motivational trajectories also differed in the shape and timing of mean changes across the four STEM courses, suggesting motivational loss trends may not be universal. Variation in trajectories also related to course performance and major persistence, with different motivational beliefs playing different roles at various times in the semester. Findings highlighted the importance of early course performance in predicting motivational change, as well as the importance of motivational changes for short- and long-term outcomes. Findings underscore the contextualized nature of motivational constructs, as both change patterns and relations to correlates differed across settings, indicating a need for researchers to carefully consider how samples are aggregated and further uncover contextual features that shape motivation. Overall, our study findings add essential understanding of motivational strengths and needs in real-world classroom contexts, with implications for the design and timing of intervention for instructors by examining what developmental patterns are common across settings and which are more particular to specific settings.

关键词： Abstraction skill   Computational thinking skill   Computer science   Computer science unplugged   SKILLS   VALIDITY  

摘要： Since computational thinking is a high-level cognitive skill and its conceptual framework has not yet been clarified, the assessment of this skill is problematic. While there is no consensus in the literature about the components of computational thinking, the abstraction component is located at a point where opinions intersect. In addition, it is predicted that computer science unplugged activities will improve computational thinking, and there is limited research examining this effect. As a result, this research aims to demonstrate the alleged effects of computer science unplugged activities on abstraction skill which is considered the critical component of computational thinking. The study, which was planned with a mixed methods research design, lasted 8 weeks. Quantitative data were obtained from the abstraction skill test developed by the researchers, while qualitative data were obtained from observations and task- based interviews after each activity. Quantitative findings showed that computer science unplugged activities significantly improve abstraction skills and thus computational thinking. Qualitative findings not only revealed that developments in the six abstracting steps caused this significant difference but also identified abstracting steps where the effect of each activity was limited. Suggestions for organizing activities to improve these steps are given. Finally, suggestions are presented for the conceptual framework of computational thinking based on abstraction.

关键词： Distributed machine learning   Parameter server   Stragglers   Heterogeneous environment   Synchronous parallel   STOCHASTIC GRADIENT DESCENT   PREDICTION  

摘要： Parameter server is widely used in distributed machine learning to accelerate training. However, the increasing heterogeneity of workers' computing capabilities leads to the issue of stragglers, making parameter synchronization challenging. To address this issue, we propose a solution called Worker-Busy Synchronous Parallel (WBSP). This approach eliminates the waiting time of fast workers during the synchronization process and decouples the gradient upload and model download of fast workers into asymmetric parts. By doing so, it allows fast workers to complete multiple steps of local training and upload more gradients to the server, improving computational resource utilization. Additionally, the global model is only updated when the slowest worker uploads the gradients, ensuring the consistency of global models that are pulled down by all workers and the convergence of the global model. Building upon WBSP, we propose an optimized version to further reduce the communication overhead. It enables parallel execution of communication and computation tasks on workers to shorten the global synchronization interval, thereby improving training speed. We conduct theoretical analyses for the proposed mechanisms. Extensive experiments verify that our mechanism can reduce the required time to achieve the target accuracy by up to 60% compared with the fastest method and increase the proportion of computation time from 55%-72% in existing methods to 91%.

关键词： Sensors   Assistive technologies   Gesture recognition   Libraries   Semantics   Computer science   Urban areas   Mobile computing   sign language translation   wearable sensing   RECOGNITION  

摘要： This paper addresses a significant problem in American Sign Language (ASL) translation systems that has been overlooked. Current designs collect excessive sensing data for each word and treat every sentence as new, requiring the collection of sensing data from scratch. This approach is time-consuming, taking hours to half a day to complete the data collection process for each user. As a result, it creates an unnecessary burden on end-users and hinders the widespread adoption of ASL systems. In this study, we identify the root cause of this issue and propose GASLA-a wearable sensor-based solution that automatically generates sentence-level sensing data from word-level data. An acceleration approach is further proposed to optimize the data generation speed. Moreover, due to the gap between the generated sentence data and directly collected sentence data, a template strategy is proposed to make the generated sentences more similar to the collected sentence. The generated data can be used to train ASL systems effectively while reducing overhead costs significantly. GASLA offers several benefits over current approaches: it reduces initial setup time and future new-sentence addition overhead; it requires only two samples per sentence compared to around ten samples in current systems; and it improves overall performance significantly.

摘要： The Chinese heroin generation is a unique yet less discussed group. Drawing upon life-history interviews with 18 former heroin drug dealers in an inland city in China, this article explores how the trajectories of the Chinese heroin generation were deeply embedded in and affected by macro social transitions, and how their collective trajectories configured the local illegal drug market. Influenced by the desire to break free from traditional constraints brought about by China’s reform and opening-up, informants voluntarily left their legitimate work to pursue self-realization but eventually became lost in the rapid changes and turned to drug trafficking. During their involvement with heroin, they adhered to mainstream Chinese cultural norms: they entered the heroin market because of “face” and “fashion,” and actively accumulated social, cultural, and criminal capital to operate their business. This article concludes that the heroin fad in China at the end of the last century is not so much a “moral crisis” as a reflection of the change in citizens' subjectivity and the unintended consequences of rapid economic development.

关键词： Computer science   Electrical and electronic engineering  

摘要： Reinforcement learning (RL) is an effective method in training dialogue policies to steer the conversation towards successful task completion. However, most RL-based methods only rely on semantic inputs that lack empathy as they ignore the user emotional information. Moreover, these methods suffer from delayed rewards caused by the user simulator returning valuable results only at dialogue end. Recently, some methods have been proposed to learn the reward function together with user emotions, but they omit considering user emotion in each dialogue turn. In this paper, we proposed an emotion-sensitive dialogue policy model (ESDP), it incorporates user emotions information into dialogue policy and selects the optimal action by the combination of top-k actions with the user emotions. The user emotion information in each turn is used as an immediate reward for the current dialogue state to solve sparse rewards and the dependency on termination. Extensive experiments validate that our method outperforms the baseline approaches when combined with different Q-Learning algorithms, and also surpasses other popular existing dialog policies' performance.