Introduction: This study investigated the role of generative Artificial Intelligence (AI) in enhancing the creative cognition of design students, examining the mediating effects of self-efficacy and anxiety reduction.Methods: A quantitative approach was employed, collecting data through online surveys from 121 design students at universities in southern China. The study utilized scales for AI knowledge and perception, self-efficacy, anxiety, and creative cognition, adapted from previous studies and evaluated on 5-point Likert scales. Data analysis was conducted using SPSS 24.0 for exploratory factor analysis and PROCESS v3.5 for mediation analysis.Results: The findings confirmed that AI positively impacted students’ innovative thinking (*β* = 0.610, *p* < 0.001). Self-efficacy (standardized *β* = 0.256, 95% CI [0.140, 0.418], *p* < 0.001) and anxiety reduction (standardized *β* = 0.093, 95% CI [0.018, 0.195], *p* < 0.05) positively mediated the relationship between generative AI and creative cognition. Additionally, a serial mediation effect through self-efficacy and anxiety reduction was observed (standardized *β* = 0.053, 95% CI [0.012, 0.114], *p* < 0.05).Discussion: Our empirical analysis demonstrates that AI positively affects design students’ innovative thinking, with self-efficacy and anxiety reduction serving as significant mediators. These findings provide valuable insights for educators and policymakers, suggesting that AI-integrated design curricula can significantly foster creative cognition, promote academic achievement, and enhance designer capabilities. Understanding AI’s impact on students’ creative processes is crucial for developing effective teaching strategies in today’s evolving educational landscape.

Erasure-coded systems require simultaneous updating of data chunks and parity chunks, thereby leading to high input/output (I/O) overhead. Following these updates, the most recent data and parity deltas are dispersed randomly in the log, which results in prolonged recovery time. This study introduces a novel approach named MPLR, which is specifically designed to expedite erasure-coded data updating and recovery. MPLR uses machine learning to categorize files into non-read-only and read-only classes, and then performs updates based on these classifications. For non-read-only files, MPLR applies in-place updates for data chunks and log-based updates for parity chunks, while setting aside disk space adjacent to the parity chunk to minimize disk seeks. For read-only files, MPLR likewise employs a mixture of in-place data updates and log-based parity updates, although without reserving space next to parity chunks. In regard to data recovery, MPLR enhances speed by accessing the parity delta stored in the allocated disk space. To further boost recovery acceleration, MPLR reads the parity delta in parallel and reduces the volume of parity delta leveraging the shareable aspects of the parity delta. Experimental results confirm the efficiency of MPLR in significantly improving update and recovery performances, compared to the state-of-the-art approaches.

Task trees are widely used as paradigms on multi-node computing environments for various computational domains, such as the sparse linear algebra and the Cholesky factorizations. These domains provides support for the computation in the Internet of Things. Meanwhile, efficient task tree scheduling approaches are the key to reducing the completion time in the environments. An algorithm for splitting a task tree into a certain number of subtrees for distributed computing on multi-node computing environments with given memory capacity is proposed. The proposed algorithm initially splits a task tree into subtree graphs, then constructs the corresponding reduction tree. The further splitting works on each subtree, which is the critical path element of the reduction tree, since the critical path is dominated by the computation and communication time. If a subtree requires more memory to process, the subtree is further split by removing the maximum node. Finally, the subtrees are merged based on the preference on the minimum increase of completion time. Extensive experimental evaluations show that the completion time can be reduced by up to 36.52% on real-world dataset, and by up to 22.92% on randomly generated instances when the proposed algorithm is applied.

In erasure-coded storage systems, updating data requires parity maintenance, which often leads to significant I/O amplification due to “write-after-read” operations. Furthermore, scattered parity placement increases disk seek overhead during repair, resulting in degraded system performance. To address these challenges, this paper proposes a Cognitive Update and Repair Method (CURM) that leverages machine learning to classify files into write-only, read-only, and read-write categories, enabling tailored update and repair strategies. For write-only and read-write files, CURM employs a data-difference mechanism combined with fine-grained I/O scheduling to minimize redundant read operations and mitigate I/O amplification. For read-write files, CURM further reserves adjacent disk space near parity blocks, supporting parallel reads and reducing disk seek overhead during repair. We implement CURM in a prototype system, Cognitive Update and Repair File System (CURFS), and conduct extensive experiments using real-world Network File System (NFS) and Microsoft Research (MSR) workloads on a 25-node cluster. Experimental results demonstrate that CURM improves data update throughput by up to 82.52%, reduces recovery time by up to 47.47%, and decreases long-term storage overhead by more than 15% compared to state-of-the-art methods including Full Logging (FL), Parity Logging (PL), Parity Logging with Reserved space (PLR), and PARIX. These results validate the effectiveness of CURM in enhancing both update and repair performance, providing a scalable and efficient solution for large-scale erasure-coded storage systems.

In erasure-coded storage systems, updating data necessitates parity updates to maintain data consistency, which leads to I/O amplification due to "write-after-read" operations. Additionally, the scattered storage of parity updates imposes significant disk seek overhead during data repair. To address these challenges, this paper proposes a Cognitive Update and Repair Method (CURM), which uses machine learning to classify files into write-only, read-only, and read-write categories, enabling customized update and repair strategies. For write-only and read-write files, CURM utilizes data difference and fine-grained I/O scheduling to reduce I/O overhead. Furthermore, CURM reserves disk space adjacent to parity chunks for read-write files, enabling efficient parallel reads and minimizing seek cost during repair. We implement CURM in a prototype storage system and evaluate its performance using real-world NFS and MSR workloads on a 25-node cluster. Experimental results show that CURM improves data update throughput by up to 82.52% and reduces data recovery time by up to 47.47%, while achieving lower storage overhead compared to state-of-the-art approaches including FL, PL, PLR, and PARIX. These results demonstrate CURM’s effectiveness in enhancing both update and recovery performance for large-scale erasure-coded storage systems.

Metadata prefetching and data placement play a critical role in enhancing access performance for file systems operating over wide-area networks. However, developing effective strategies for metadata prefetching in environments with concurrent workloads and for data placement across distributed networks remains a significant challenge. This study introduces novel and efficient methodologies for metadata prefetching and data placement, leveraging fine-grained control of prefetching strategies and variable-sized data fragment writing to optimize the I/O bandwidth of distributed file systems. The proposed metadata prefetching technique employs dynamic workload analysis to identify dominant workload patterns and adaptively refines prefetching policies, thereby boosting metadata access efficiency under concurrent scenarios. Meanwhile, the data placement strategy improves write performance by storing data fragments locally within the nearest data center and transmitting only the fragment location metadata to the remote data center hosting the original file. Experimental evaluations using real-world system traces demonstrate that the proposed approaches reduce metadata access times by up to 33.5% and application data access times by 17.19% compared to state-of-the-art techniques.