Knowledge graph representation learning with simplifying hierarchical feature propagation

Graph neural networks (GNN) have emerged as a new state-of-the-art for learning knowledge graph representations. Although they have shown impressive performance in recent studies, how to efficiently and effectively aggregate neighboring features is not well designed. To tackle this challenge, we propose the simplifying heterogeneous graph neural network (SHGNet), a generic framework that discards the two standard operations in GNN, including the transformation matrix and nonlinear activation. SHGNet, in particular, adopts only the essential component of neighborhood aggregation in GNN and incorporates relation features into feature propagation. Furthermore, to capture complex structures, SHGNet utilizes a hierarchical aggregation architecture, including node aggregation and relation weighting. Thus, the proposed model can treat each relation differently and selectively aggregate informative features. SHGNet has been evaluated for link prediction tasks on three real-world benchmark datasets. The experimental results show that SHGNet significantly promotes efficiency while maintaining superior performance, outperforming all the existing models in 3 out of 4 metrics on NELL-995 and in 4 out of 4 metrics on FB15k-237 dataset.     

高校体育文化建设中信息环境的构建与变迁

高校体育文化,作为一种带有明确范围限定和内容限定的亚文化形态,其建设和发展的实质,是内涵中体育信息环境的构建过程。随着高校体育文化中体育信息环境的内容、来源与群体层面都发生了变迁,高校体育文化在文化循环的过程中,出现了文化表征方式从行为向非行为的分化,并表现为高校体育文化信息环境变迁对体育活动参与者的行为阻滞作用。     

整数对的2-Ary自适应窗口表示算法

在有些公钥密码体制的实现中,特别是大多数数字签名（除了RSA）的验证过程,不仅需要指数运算,还需要能直接计算指数对。本文研究分析了整数对的从左到右的2阶自适应窗口表示（2-AWR）,使用该表示可以使得Shamir算法实时实现,并且在同类算法中具有明显的速度优势。     

量子力学表象与群表示中线性代数方法的应用

对表象变换,群表示与线性代数中n维矢量空间中坐标变换的相似性进行了类比分析,结果表明：线性代数是二者的数学基础。物理规律对应着确定的群,群的不可约表示对H空间的态矢分类可方便地决定一个量子系统的结构。并对H2O的电子态进行了分析。     

流数据聚类研究综述

流数据挖掘技术是数据挖掘领域的新研究方向之一,而聚类研究又是其重要的内容。本文介绍了流数据基本特点,在统一流聚类表示模型的基础上,对现有流数据聚类算法进行了总结,并进一步提出了流数据聚类技术的研究方向和前景。     

Despicable Others: Animated Othering as Equipment for Living in the Era of Trump

This essay comparatively examines U.S.-Russian and U.S.-Mexican interculturality in the Despicable Me movie franchise. We argue that cultural transformations of the main protagonists – Russian Gru and Mexican El Macho – are politically significant cases of U.S.-Russian and U.S.-Mexican interculturality, reflected and constructed by popular culture in general, and by animated cartoons in particular. Conceptualizing and analyzing animation as “equipment for living,” we demonstrate how these cases serve as epistemological and pedagogical tools of confirming U.S.-dominant ideology, and discursively and visually constructing (Russian and Mexican) Others the way that they are currently understood and treated in (the non-fictional) U.S. society.     

体育教学中提高学生内部反馈能力的研究

文章对信息反馈系统中内部反馈和外部反馈的不同作用以及对它们之间的相互关系的研究,内部反馈在动作技能形成中起着重要作用。在体育教学训练实践中应采取：建立动作表象和动作概念;进行表象训练;向练习者提供正确动作的感觉信息;利用动作间的迁移和组织练习者自评自测的方式提高练习者的内部反馈能力。     

Empowering teacher voices in an education policy discussion: Paradoxes of representation

This article investigates the problems and paradoxes of attempting to empower teacher voices within the context of an international conference of policy-makers, academic researchers and practitioners. We examine the distribution of talk within a teacher workshop: who spoke, how, and to whom did the group and broader audiences listen? We trace the emergence of ideas in the workshop discussions and their trajectory into the joint teacher-policy-maker panel in the conference and in the post-conference summary report. We identify four factors shaping the realisation of teacher voice – repertoires, social position, topics and gatekeepers – and highlight paradoxes of teacher representation.     

基于文献的知识表示

知识表示是人工智能研究中的基本问题之一,也是网络时代知识服务的基础。文章描述了一种新的适用于数字图书馆的知识表示方法,在此基础上探讨了知识库的特性和推理机制,并研究了数字图书馆中提供知识服务的模式。     

A Critique of Internet Polls as Symbolic Representation and Pseudo-Events

The Internet continues to grow as an information and entertainment medium. Internet growth has implications for the news industry. Twenty-four hour news networks such as CNN and MSNBC regularly encourage viewers of their television programs to visit their Web sites. While visiting news Web sites, visitors are invited to participate in opinion polls. Unfortunately, these online opinion polls are not scientific and have little real news value. In spite of these limitations, news Web sites' Internet polls are often treated as serious topics in broadcast news discussions. This article examines media organizations' Internet online polls and critiques them as instances of symbolic representation and pseudo-events that have arisen largely out of the integration of print, broadcast, and Internet media.