随着大数据时代的到来,如何从多源异构数据中准确地识别网络安全实体是构建网络安全知识图谱的基础问题。因此本文针对网络安全相关文本数据,研究支持海量网络数据的安全实体识别算法,为构建网络安全知识图谱奠定基础。针对海量的文本...随着大数据时代的到来,如何从多源异构数据中准确地识别网络安全实体是构建网络安全知识图谱的基础问题。因此本文针对网络安全相关文本数据,研究支持海量网络数据的安全实体识别算法,为构建网络安全知识图谱奠定基础。针对海量的文本类网络数据中安全实体的高效精准抽取问题,本文基于Hadoop分布式计算框架提出改进的条件随机场(conditional random fields,CRF)算法,对数据集进行有效分割,实现安全实体的高效准确识别。在大规模真实网络数据集上的实验证明,本文提出的算法达到了较高的网络安全实体识别准确率,同时提高了识别的效率。展开更多
In this work, the damage and penetration behavior of aluminum foam at various types of impact were examined through experiments. The impact energy of a striker was applied on the fixed aluminum foam having a thickness...In this work, the damage and penetration behavior of aluminum foam at various types of impact were examined through experiments. The impact energy of a striker was applied on the fixed aluminum foam having a thickness of 25 mm while increasing its impact by 2 J at each strike from 6 J to 16 J. The results show that the impact energies from 6 J to 12 J could not penetrate aluminum foam. However, the aluminum foam applied with the impact energy of 12 J incurred severe damages on its lower part. Finally, the aluminum foam applied with the impact energy of 14 J was penetrated. The striker having the impact energy of 6 J could penetrate aluminum foam around 10 mm. At this moment, aluminum foam could absorb the impact energy of around 9 J. When the impact energy of 14 J was applied on the aluminum foam, the aluminum foam was penetrated and it absorbed the impact energy of around 17.2 J. It is possible to create the safer structure against impact using the results of this work. The simulation results for the verification of the experimental results imply that the results for all the experiments in this work are reliable. It is possible to predict the structural safety of the aluminum foam for an impact if the impact behavior of aluminum foam performed in this work is utilized.展开更多
文摘随着大数据时代的到来,如何从多源异构数据中准确地识别网络安全实体是构建网络安全知识图谱的基础问题。因此本文针对网络安全相关文本数据,研究支持海量网络数据的安全实体识别算法,为构建网络安全知识图谱奠定基础。针对海量的文本类网络数据中安全实体的高效精准抽取问题,本文基于Hadoop分布式计算框架提出改进的条件随机场(conditional random fields,CRF)算法,对数据集进行有效分割,实现安全实体的高效准确识别。在大规模真实网络数据集上的实验证明,本文提出的算法达到了较高的网络安全实体识别准确率,同时提高了识别的效率。
基金Project(2011-0006548)supported by the Basic Research Program through the National Research Foundation of Korea
文摘In this work, the damage and penetration behavior of aluminum foam at various types of impact were examined through experiments. The impact energy of a striker was applied on the fixed aluminum foam having a thickness of 25 mm while increasing its impact by 2 J at each strike from 6 J to 16 J. The results show that the impact energies from 6 J to 12 J could not penetrate aluminum foam. However, the aluminum foam applied with the impact energy of 12 J incurred severe damages on its lower part. Finally, the aluminum foam applied with the impact energy of 14 J was penetrated. The striker having the impact energy of 6 J could penetrate aluminum foam around 10 mm. At this moment, aluminum foam could absorb the impact energy of around 9 J. When the impact energy of 14 J was applied on the aluminum foam, the aluminum foam was penetrated and it absorbed the impact energy of around 17.2 J. It is possible to create the safer structure against impact using the results of this work. The simulation results for the verification of the experimental results imply that the results for all the experiments in this work are reliable. It is possible to predict the structural safety of the aluminum foam for an impact if the impact behavior of aluminum foam performed in this work is utilized.
