Recently, the emergence of pre-trained models(PTMs) has brought natural language processing(NLP) to a new era. In this survey, we provide a comprehensive review of PTMs for NLP. We first briefly introduce language rep...Recently, the emergence of pre-trained models(PTMs) has brought natural language processing(NLP) to a new era. In this survey, we provide a comprehensive review of PTMs for NLP. We first briefly introduce language representation learning and its research progress. Then we systematically categorize existing PTMs based on a taxonomy from four different perspectives. Next,we describe how to adapt the knowledge of PTMs to downstream tasks. Finally, we outline some potential directions of PTMs for future research. This survey is purposed to be a hands-on guide for understanding, using, and developing PTMs for various NLP tasks.展开更多
访问控制是网络安全的基础技术。随着大数据技术与开放式网络的发展,互联网用户的访问行为变得越来越灵活。传统的访问控制机制主要从规则自动生成和规则匹配优化两方面来提升访问控制的工作效率,大多采用遍历匹配机制,存在计算量大、...访问控制是网络安全的基础技术。随着大数据技术与开放式网络的发展,互联网用户的访问行为变得越来越灵活。传统的访问控制机制主要从规则自动生成和规则匹配优化两方面来提升访问控制的工作效率,大多采用遍历匹配机制,存在计算量大、效率低等问题,难以满足开放式环境下访问控制动态、高效的需求。受人工智能领域中的分布式嵌入技术的启发,提出一种基于向量表征与计算的访问控制的VRCAC(Vector Representation and Computation based Access Control)模型。首先将访问控制规则转化为数值型向量,使得计算机能够以数值计算的方式实现快速的访问判定,用户向量与权限向量的位置关系可用两者的内积值表示,通过比较内积值与关系阈值,可以快速判断用户与权限的关系。此方法降低了访问控制执行的时间复杂度,从而提高了开放式大数据环境下的访问控制的执行效率。最后在两个真实数据集上,采用准确率、误报率等多种评价指标进行了比较实验,验证了所提方法的有效性。展开更多
Edge Computing is one of the radically evolving systems through generations as it is able to effectively meet the data saving standards of consumers,providers and the workers. Requisition for Edge Computing based ite...Edge Computing is one of the radically evolving systems through generations as it is able to effectively meet the data saving standards of consumers,providers and the workers. Requisition for Edge Computing based items havebeen increasing tremendously. Apart from the advantages it holds, there remainlots of objections and restrictions, which hinders it from accomplishing the needof consumers all around the world. Some of the limitations are constraints oncomputing and hardware, functions and accessibility, remote administration andconnectivity. There is also a backlog in security due to its inability to create a trustbetween devices involved in encryption and decryption. This is because securityof data greatly depends upon faster encryption and decryption in order to transferit. In addition, its devices are considerably exposed to side channel attacks,including Power Analysis attacks that are capable of overturning the process.Constrained space and the ability of it is one of the most challenging tasks. Toprevail over from this issue we are proposing a Cryptographic LightweightEncryption Algorithm with Dimensionality Reduction in Edge Computing. Thet-Distributed Stochastic Neighbor Embedding is one of the efficient dimensionality reduction technique that greatly decreases the size of the non-linear data. Thethree dimensional image data obtained from the system, which are connected withit, are dimensionally reduced, and then lightweight encryption algorithm isemployed. Hence, the security backlog can be solved effectively using thismethod.展开更多
In order to improve the light welfare of Nile tilapia in aquaculture,the influence of hunger level on light spectrum preference of Nile tilapia was explored in this study.The whole experiment was based on the emptying...In order to improve the light welfare of Nile tilapia in aquaculture,the influence of hunger level on light spectrum preference of Nile tilapia was explored in this study.The whole experiment was based on the emptying of the gastrointestinal contents,and carried out under the controlled laboratory conditions.The light spectrum preference was assessed by counting the head location of fish in each experimental tank,which containing seven compartments(i.e.,red,blue,white,yellow,black,green and public area).t-Distributed Stochastic Neighbor Embedding(t-SNE)was adopted to visualize the hunger level-based dynamic preference on light spectrum in two-dimensional space.According to the clustering results,significant differences in light spectrum preferences of Nile tilapia,under the different hunger levels,were indicated.In addition,the average visit frequency in green compartment was significantly lower than that in other color compartments throughout the whole experiment,and the total visit frequency in red compartment was relatively higher during the whole experiment.展开更多
为了提取被强噪声淹没的机械设备振动信号中蕴含的微弱故障特征,依据有用信号和噪声在空间分布特性的不同,将流形学习的方法引入到信号降噪中,提出一种将双树复小波包(DTCWPT)和t分布随机近邻嵌入(t-SNE)结合的去噪方法,充分利用了DTCWP...为了提取被强噪声淹没的机械设备振动信号中蕴含的微弱故障特征,依据有用信号和噪声在空间分布特性的不同,将流形学习的方法引入到信号降噪中,提出一种将双树复小波包(DTCWPT)和t分布随机近邻嵌入(t-SNE)结合的去噪方法,充分利用了DTCWPT分解的多尺度特性以及t-SNE的非线性降维能力。将振动信号进行双树复小波包分解,依据各尺度小波包系数Shannon熵值搜索最佳小波包基,利用提出的新的阈值函数,对最佳小波包基的小波包系数进行去噪并单支重构组成高维信号空间,然后,采用t-SNE提取高维空间的低维流形,对低维信号序列进一步采用阈值去噪,利用谱回归分析重构回一维信号序列。最后,通过对仿真信号与滚动轴承振动信号进行去噪,结果证实了方法具有良好的非线性去噪性能,将仿真信号的信噪比从-1提高到8.6 d B,并且能更有效的提取强噪声干扰下滚动轴承的故障特征频率。展开更多
Knowledge graph representation has been a long standing goal of artificial intelligence. In this paper,we consider a method for knowledge graph embedding of hyper-relational data, which are commonly found in knowledge...Knowledge graph representation has been a long standing goal of artificial intelligence. In this paper,we consider a method for knowledge graph embedding of hyper-relational data, which are commonly found in knowledge graphs. Previous models such as Trans（E, H, R） and CTrans R are either insufficient for embedding hyper-relational data or focus on projecting an entity into multiple embeddings, which might not be effective for generalization nor accurately reflect real knowledge. To overcome these issues, we propose the novel model Trans HR, which transforms the hyper-relations in a pair of entities into an individual vector, serving as a translation between them. We experimentally evaluate our model on two typical tasks—link prediction and triple classification.The results demonstrate that Trans HR significantly outperforms Trans（E, H, R） and CTrans R, especially for hyperrelational data.展开更多
基金the National Natural Science Foundation of China(Grant Nos.61751201 and 61672162)the Shanghai Municipal Science and Technology Major Project(Grant No.2018SHZDZX01)and ZJLab。
文摘Recently, the emergence of pre-trained models(PTMs) has brought natural language processing(NLP) to a new era. In this survey, we provide a comprehensive review of PTMs for NLP. We first briefly introduce language representation learning and its research progress. Then we systematically categorize existing PTMs based on a taxonomy from four different perspectives. Next,we describe how to adapt the knowledge of PTMs to downstream tasks. Finally, we outline some potential directions of PTMs for future research. This survey is purposed to be a hands-on guide for understanding, using, and developing PTMs for various NLP tasks.
文摘访问控制是网络安全的基础技术。随着大数据技术与开放式网络的发展,互联网用户的访问行为变得越来越灵活。传统的访问控制机制主要从规则自动生成和规则匹配优化两方面来提升访问控制的工作效率,大多采用遍历匹配机制,存在计算量大、效率低等问题,难以满足开放式环境下访问控制动态、高效的需求。受人工智能领域中的分布式嵌入技术的启发,提出一种基于向量表征与计算的访问控制的VRCAC(Vector Representation and Computation based Access Control)模型。首先将访问控制规则转化为数值型向量,使得计算机能够以数值计算的方式实现快速的访问判定,用户向量与权限向量的位置关系可用两者的内积值表示,通过比较内积值与关系阈值,可以快速判断用户与权限的关系。此方法降低了访问控制执行的时间复杂度,从而提高了开放式大数据环境下的访问控制的执行效率。最后在两个真实数据集上,采用准确率、误报率等多种评价指标进行了比较实验,验证了所提方法的有效性。
文摘Edge Computing is one of the radically evolving systems through generations as it is able to effectively meet the data saving standards of consumers,providers and the workers. Requisition for Edge Computing based items havebeen increasing tremendously. Apart from the advantages it holds, there remainlots of objections and restrictions, which hinders it from accomplishing the needof consumers all around the world. Some of the limitations are constraints oncomputing and hardware, functions and accessibility, remote administration andconnectivity. There is also a backlog in security due to its inability to create a trustbetween devices involved in encryption and decryption. This is because securityof data greatly depends upon faster encryption and decryption in order to transferit. In addition, its devices are considerably exposed to side channel attacks,including Power Analysis attacks that are capable of overturning the process.Constrained space and the ability of it is one of the most challenging tasks. Toprevail over from this issue we are proposing a Cryptographic LightweightEncryption Algorithm with Dimensionality Reduction in Edge Computing. Thet-Distributed Stochastic Neighbor Embedding is one of the efficient dimensionality reduction technique that greatly decreases the size of the non-linear data. Thethree dimensional image data obtained from the system, which are connected withit, are dimensionally reduced, and then lightweight encryption algorithm isemployed. Hence, the security backlog can be solved effectively using thismethod.
基金supported by the National Key R&D Program of China(Grant No.2017YFB0404000)the Key R&D Program of Ningxia Hui Autonomous Region(Grant No.2018BBF02009)Open Fund of Yunnan Province Key Laboratory of Food Processing and Safety Control(Grant No.K16-507106-007)。
文摘In order to improve the light welfare of Nile tilapia in aquaculture,the influence of hunger level on light spectrum preference of Nile tilapia was explored in this study.The whole experiment was based on the emptying of the gastrointestinal contents,and carried out under the controlled laboratory conditions.The light spectrum preference was assessed by counting the head location of fish in each experimental tank,which containing seven compartments(i.e.,red,blue,white,yellow,black,green and public area).t-Distributed Stochastic Neighbor Embedding(t-SNE)was adopted to visualize the hunger level-based dynamic preference on light spectrum in two-dimensional space.According to the clustering results,significant differences in light spectrum preferences of Nile tilapia,under the different hunger levels,were indicated.In addition,the average visit frequency in green compartment was significantly lower than that in other color compartments throughout the whole experiment,and the total visit frequency in red compartment was relatively higher during the whole experiment.
文摘为了提取被强噪声淹没的机械设备振动信号中蕴含的微弱故障特征,依据有用信号和噪声在空间分布特性的不同,将流形学习的方法引入到信号降噪中,提出一种将双树复小波包(DTCWPT)和t分布随机近邻嵌入(t-SNE)结合的去噪方法,充分利用了DTCWPT分解的多尺度特性以及t-SNE的非线性降维能力。将振动信号进行双树复小波包分解,依据各尺度小波包系数Shannon熵值搜索最佳小波包基,利用提出的新的阈值函数,对最佳小波包基的小波包系数进行去噪并单支重构组成高维信号空间,然后,采用t-SNE提取高维空间的低维流形,对低维信号序列进一步采用阈值去噪,利用谱回归分析重构回一维信号序列。最后,通过对仿真信号与滚动轴承振动信号进行去噪,结果证实了方法具有良好的非线性去噪性能,将仿真信号的信噪比从-1提高到8.6 d B,并且能更有效的提取强噪声干扰下滚动轴承的故障特征频率。
基金partially supported by the National Natural Science Foundation of China(Nos.61302077,61520106007,61421061,and 61602048)
文摘Knowledge graph representation has been a long standing goal of artificial intelligence. In this paper,we consider a method for knowledge graph embedding of hyper-relational data, which are commonly found in knowledge graphs. Previous models such as Trans（E, H, R） and CTrans R are either insufficient for embedding hyper-relational data or focus on projecting an entity into multiple embeddings, which might not be effective for generalization nor accurately reflect real knowledge. To overcome these issues, we propose the novel model Trans HR, which transforms the hyper-relations in a pair of entities into an individual vector, serving as a translation between them. We experimentally evaluate our model on two typical tasks—link prediction and triple classification.The results demonstrate that Trans HR significantly outperforms Trans（E, H, R） and CTrans R, especially for hyperrelational data.
