Abstract
Deep learning has brought a dramatic development in molecular property prediction that is crucial in the field of drug discovery using various representations such as fingerprints, SMILES, ...and graphs. In particular, SMILES is used in various deep learning models via character-based approaches. However, SMILES has a limitation in that it is hard to reflect chemical properties. In this paper, we propose a new self-supervised method to learn SMILES and chemical contexts of molecules simultaneously in pre-training the Transformer. The key of our model is learning structures with adjacency matrix embedding and learning logics that can infer descriptors via Quantitative Estimation of Drug-likeness prediction in pre-training. As a result, our method improves the generalization of the data and achieves the best average performance by benchmarking downstream tasks. Moreover, we develop a web-based fine-tuning service to utilize our model on various tasks.
Analysis of multi-physics systems and the visualization of simulation data are crucial and difficult in computational science and engineering. In Korea, Korea Institute of Science and Technology ...Information KISTI developed EDISON, a web-based computational science simulation platform, and it is now the ninth year since the service started. Hitherto, the EDISON platform has focused on providing a robust simulation environment and various computational science analysis tools. However, owing to the increasing issues in collaborative research, data format standardization has become more important. In addition, as the visualization of simulation data becomes more important for users to understand, the necessity of analyzing input / output data information for each software is increased. Therefore, it is necessary to organize the data format and metadata for the representative software provided by EDISON. In this paper, we analyzed computational fluid dynamics (CFD) and computational structural dynamics (CSD) simulation software in the field of mechanical engineering where several physical phenomena (fluids, solids, etc.) are complex. Additionally, in order to visualize various simulation result data, we used existing web visualization tools developed by third parties. In conclusion, based on the analysis of these data formats, it is possible to provide a foundation of multi-physics and a web-based visualization environment, which will enable users to focus on simulation more conveniently. Keywords: computational science and engineering, EDISON, data format, metadata, visualization environment A preliminary version of this paper was presented at APIC-IST 2019, and was selected as a best paper. This research was supported by the KISTI Program (No. K-19-L02-C05-S01), the EDISON Program through the National Research Foundation of Korea (NRF) (No. NRF-2011-0020576). A Grant was awarded by the Ministry of Science and ICT (MSIT) under the Program for Returners for R&D.
Analysis of multi-physics systems and the visualization of simulation data are crucial and difficult in computational science and engineering. In Korea, Korea Institute of Science and Technology ...Information KISTI developed EDISON, a web-based computational science simulation platform, and it is now the ninth year since the service started. Hitherto, the EDISON platform has focused on providing a robust simulation environment and various computational science analysis tools. However, owing to the increasing issues in collaborative research, data format standardization has become more important. In addition, as the visualization of simulation data becomes more important for users to understand, the necessity of analyzing input / output data information for each software is increased. Therefore, it is necessary to organize the data format and metadata for the representative software provided by EDISON. In this paper, we analyzed computational fluid dynamics (CFD) and computational structural dynamics (CSD) simulation software in the field of mechanical engineering where several physical phenomena (fluids, solids, etc.) are complex. Additionally, in order to visualize various simulation result data, we used existing web visualization tools developed by third parties. In conclusion, based on the analysis of these data formats, it is possible to provide a foundation of multi-physics and a web-based visualization environment, which will enable users to focus on simulation more conveniently.
The computational science and engineering field, which combines basic science and computing technology, has emerged as a third scientific methodology, following theories and experiments. This study ...aimed to identify factors and relationships that affect the continued use of the computational science and engineering (CSE) platform for its successful operation, utilization, and diffusion. To that end, the quality factors of the platform were derived by combining the information system success model and the technology acceptance model. These factors affected user satisfaction and intention to reuse through users’ perceived usefulness and perceived ease of use of the platform. An empirical analysis was conducted through a questionnaire survey of 373 users of the EDISON platform, a representative CSE platform in Korea. The results revealed that all quality factors have a positive influence on perceived usefulness and perceived ease of use. Specifically, information quality has a significant influence on perceived ease of use, and system quality has a significant influence on perceived usefulness. Perceived ease of use has a greater impact on user satisfaction than perceived usefulness, and satisfaction affects intention to reuse. The results can contribute to the development of CSE platforms and the development strategy to expand the number of users.
The computational science and engineering field, which combines basic science and computing technology, has emerged as a third scientific methodology, following theories and experiments. This study ...aimed to identify factors and relationships that affect the continued use of the computational science and engineering (CSE) platform for its successful operation, utilization, and diffusion. To that end, the quality factors of the platform were derived by combining the information system success model and the technology acceptance model. These factors affected user satisfaction and intention to reuse through users' perceived usefulness and perceived ease of use of the platform. An empirical analysis was conducted through a questionnaire survey of 373 users of the EDISON platform, a representative CSE platform in Korea. The results revealed that all quality factors have a positive influence on perceived usefulness and perceived ease of use. Specifically, information quality has a significant influence on perceived ease of use, and system quality has a significant influence on perceived usefulness. Perceived ease of use has a greater impact on user satisfaction than perceived usefulness, and satisfaction affects intention to reuse. The results can contribute to the development of CSE platforms and the development strategy to expand the number of users. Keywords: Computational Science and Engineering Platform, EDISON Platform, Information Systems Success Model, Technology Acceptance Model, Structural Equation Analysis
Composite Measures of Supercomputer Technology Kim, Nam-Gyu; On, Noo Ri; Koh, Myoung-Ju ...
KSII transactions on Internet and information systems,
08/2019, Letnik:
13, Številka:
8
Journal Article
Recenzirano
Odprti dostop
We have developed composite measures of supercomputer technology, reflecting various factors of supercomputers using Martino’s scoring model. CPUs, accelerators, memory, interconnection networks, and ...power consumption are chosen as factors of the model. The weight values of the factors are derived based on a survey of 129 domestic and international experts. The measured values are then standardized to integrate measurement units of the factors in the model. This model has been applied to 50 supercomputers, and rank correlation analysis was performed using representative measures. As a consequence, the ranking drastically changes except for the 1st and 2nd supercomputers on the TOP500. In addition, the characteristics of memory and interconnection networks influence the ranking, and the results demonstrate that the proposed model has low correlations with HPL and HPCG but a high correlation with Green500. This indicates that power consumption is an important factor that has a significant effect on the measures of supercomputer technology. In addition, it is determined that the differences between the HPL ranking and the proposed model ranking are influenced by power consumption, CPU theoretical peak performance, and main memory bandwidth in order of significance. In conclusion, the composite measures proposed in this study are more suitable for comprehensively describing supercomputer technology than existing performance measures. The findings of this study are expected to support decision making related to management and policy in the procurement and operation of supercomputers.
We have developed an information management tool for the EDISON (EDucation-research Integration through Simulation On the Net) open platform. EDISON is, at present, a web-based simulation service for ...education and research in five computational areas, namely, nanophysics, fluid dynamics, chemistry, structural dynamics, and computer aided optimal design. The EDISON open platform consists of three tiers: EDISON application framework, EDISON middleware, and EDISON infra-resources. The platform provides web portals for education and research in areas such as computational fluid dynamics, computational chemistry, computational nanophysics, computational structural dynamics, and computer aided optimal design along with user service.
The main purpose of this research is to test the behavior of the release version of the EDISON Open-Platform under normal operating conditions. This management tool has been implemented using the RESTful API designed in EDISON middleware. The intention is to check co-operation between the middleware and the infrastructure. Suggested tools include User management, Simulation and Job management, and Simulation software (i.e., solver) testing. Finally, it is considered meaningful to develop a management tool that is not supported in other web-based online simulation services.
A new capability of our well-known NEMO 3-D simulator (Ref. Klimeck et al., 2007 10) is introduced by carefully investigating the utility of III–V semiconductor quantum dots as infrared ...photodetectors at a wavelength of 1.2–1.5 μm. We not only present a detailed description of the simulation methodology coupled to the atomistic sp3d5s∗ tight-binding band model, but also validate the suggested methodology with a focus on a proof of principle on small GaAs quantum dots (QDs). Then, we move the simulation scope to optical properties of realistically sized dome-shaped InAs/GaAs QDs that are grown by self-assembly and typically contain a few million atoms. Performing numerical experiments with a variation in QD size, we not only show that the strength of ground state inter-band light transitions can be optimized via QD size-engineering, but also find that the hole ground state wavefunction serves as a control factor of transition strengths. Finally, we briefly introduce the web-based cyber infrastructure that is developed as a government-funded project to support online education and research via TCAD simulations. This work not only serves as a useful guideline to experimentalists for potential device designs and other modelers for the self-development of optical TCAD, but also provides a good chance to learn about the science gateway project ongoing in the Republic of Korea.
In the 21 st century, IT reform has led to the development of cyber-infrastructure owing to the outstanding enhancement of computer and network performance. The ripple effect has continued to ...increase. Accordingly, this study suggests a new computational research environment using mobile devices. In order to simplify the access of supercomputer, Science AppStore, task management and virtualization technologies are developed on mobile devices. User can be able to research by utilizing computational science SW such as compressible flow solver and nano device simulation tool that in installed on supercomputer in mobile environments. Also, this research environment makes it possible to monitor the simulation result and covers 14 university, 33 subjects, and 1,202 individuals.