基于融合元路径权重的异质网络表征学习

针对基于单条元路径的异质网络表征缺失异质信息网络中结构信息及其它元路径语义信息的问题,本文提出了基于融合元路径权重的异质网络表征学习方法.该方法对异质信息网络中元路径集合进行权重学习,进而对基于不同元路径的低维表征进行加权融合,得到融合不同元路径语义信息的异质网络表征.实验结果表明,基于融合元路径权重的异质网络表征学习具有良好的表征学习能力,可有效应用于数据挖掘.     

An eye movement analysis of highlighting and graphic organizer study aids for learning from expository text

This study uses eye tracking technology to examine how study aids such as highlighting and graphic organizers affect cognitive processing during learning. Participants were 130 college students randomly assigned to one of five experimental conditions. In the control group, students read a plain text; in two behaviorally passive conditions, students read a text with key words colored in red or read the same text along with a filled-in graphic organizer; and in two behaviorally active conditions, students either highlighted key words in a text or filled in an empty graphic organizer. Students took tests of rote memory (cloze test) and comprehension (summary test). Asking students to fill in a graphic organizer or providing a filled-in graphic organizer resulted in improvements in performance on both tests, whereas asking students to highlight the text or providing highlighted text improved performance only in the rote memory test compared to students who did not receive any study aids. Eye tracking measures showed that highlighting (in both conditions) primed the cognitive process of selecting: students spent more time fixating on those words colored in red compared with the control condition. In contrast, eye tracking measures showed that graphic organizers (in both conditions) primed the cognitive processes of selecting, organizing and integrating since the inclusion of an organizer substantially affected both where their eyes fixated and moved (i.e. transitions) within the text.     

复杂异构数据的表征学习综述

随着智能时代和大数据时代的到来,各种复杂异构数据不断涌现,成为数据驱动的人工智能方法、机器学习模型的基础。复杂异构数据的表征直接关系着后续模型的学习性能,因此如何有效地表征复杂异构数据成为机器学习的一个重要研究领域。文中首先介绍了数据表征的多种类型,并提出了现有数据表征方法面临的挑战;其次,根据数据类型将数据划分成单一类型数据和复合类型数据,针对单一类型数据,分别介绍了4种典型数据的表征学习发展现状和代表算法,包含离散数据、网络数据、文本数据和图像数据;然后,详细介绍了4种由多个单一数据或数据源复合而成的复杂数据,包含了离散特征与连续特征混合的结构化数据、属性数据与复杂网络复合的属性网络数据、来自不同领域的跨领域数据和由多种数据类型复合的多模态数据,分别介绍了基于上述复杂数据的表征学习现状以及最新的表征学习模型;最后,对复杂异构数据表征学习的发展趋势进行了探讨。     

Social Representations of High School Students About Mathematics Assessment

The perceptions of students about assessment in mathematics classes have been sparsely investigated. In order to fill this gap, this qualitative study aims to identify the social representations (understood as the system of values, ideas, and practices about a social object) of high school students regarding assessment in mathematics. We used semistructured focus group interviews to obtain data (N = 50). The data were examined using theoretical thematic analysis with the assistance of the specialized ATLAS.ti software. Eight themes or social representations emerged (e.g., assessment that measures the students' learning or assessment that measures the advances or what has been acquired). The results are consistent with literature that shows that students' representations about assessment in mathematics are closely linked to their representations of mathematics itself and the learning of it and also to their representation of what is fair and what is not.     

HiWalk: Learning node embeddings from heterogeneous networks

Heterogeneous networks, such as bibliographical networks and online business networks, are ubiquitous in everyday life. Nevertheless, analyzing them for high-level semantic understanding still poses a great challenge for modern information systems. In this paper, we propose HiWalk to learn distributed vector representations of the nodes in heterogeneous networks. HiWalk is inspired by the state-of-the-art representation learning algorithms employed in the context of both homogeneous networks and heterogeneous networks, based on word embedding learning models. Different from existing methods in the literature, the purpose of HiWalk is to learn vector representations of the targeted set of nodes by leveraging the other nodes as “background knowledge”, which maximizes the structural correlations of contiguous nodes. HiWalk decomposes the adjacent probabilities of the nodes and adopts a hierarchical random walk strategy, which makes it more effective, efficient and concentrated when applied to practical large-scale heterogeneous networks. HiWalk can be widely applied in heterogeneous networks environments to analyze targeted types of nodes. We further validate the effectiveness of the proposed HiWalk through multiple tasks conducted on two real-world datasets.     

Text-enhanced network representation learning

Network representation learning called NRL for short aims at embedding various networks into lowdimensional continuous distributed vector spaces. Most existing representation learning methods focus on learning representations purely based on the network topology, i.e., the linkage relationships between network nodes, but the nodes in lots of networks may contain rich text features, which are beneficial to network analysis tasks, such as node classification, link prediction and so on. In this paper, we propose a novel network representation learning model, which is named as Text-Enhanced Network Representation Learning called TENR for short, by introducing text features of the nodes to learn more discriminative network representations, which come from joint learning of both the network topology and text features, and include common influencing factors of both parties. In the experiments, we evaluate our proposed method and other baseline methods on the task of node classification. The experimental results demonstrate that our method outperforms other baseline methods on three real-world datasets.     

Improving collaborative learning in the classroom: Text mining based grouping and representing

Orchestrating collaborative learning in the classroom involves tasks such as forming learning groups with heterogeneous knowledge and making learners aware of the knowledge differences. However, gathering information on which the formation of appropriate groups and the creation of graphical knowledge representations can be based is very effortful for teachers. Tools supporting cognitive group awareness provide such representations to guide students during their collaboration, but mainly rely on specifically created input. Our work is guided by the questions of how the analysis and visualization of cognitive information can be supported by automatic mechanisms (especially using text mining), and what effects a corresponding tool can achieve in the classroom. We systematically compared different methods to be used in a Grouping and Representing Tool (GRT), and evaluated the tool in an experimental field study. Latent Dirichlet Allocation proved successful in transforming the topics of texts into values as a basis for representing cognitive information graphically. The Vector Space Model with Euclidian distance based clustering proved to be particularly well suited for detecting text differences as a basis for group formation. The subsequent evaluation of the GRT with 54 high school students further confirmed the GRT’s impact on learning support: students who used the tool added twice as many concepts in an essay after discussing as those in the unsupported group. These results show the potential of the GRT to support both teachers and students.     

Comparative Analysis of Text Representation Methods Using Classification

In our work, we review and empirically evaluate five different raw methods of text representation that allow automatic processing of Wikipedia articles. The main contribution of the article—evaluation of approaches to text representation for machine learning tasks—indicates that the text representation is fundamental for achieving good categorization results. The analysis of the representation methods creates a baseline that cannot be compensated for even by sophisticated machine learning algorithms. It confirms the thesis that proper data representation is a prerequisite for achieving high-quality results of data analysis. Evaluation of the text representations was performed within the Wikipedia repository by examination of classification parameters observed during automatic reconstruction of human-made categories. For that purpose, we use a classifier based on a support vector machines method, extended with multilabel and multiclass functionalities. During classifier construction we observed parameters such as learning time, representation size, and classification quality that allow us to draw conclusions about text representations. For the experiments presented in the article, we use data sets created from Wikipedia dumps. We describe our software, called Matrix’u, which allows a user to build computational representations of Wikipedia articles. The software is the second contribution of our research, because it is a universal tool for converting Wikipedia from a human-readable form to a form that can be processed by a machine. Results generated using Matrix’u can be used in a wide range of applications that involve usage of Wikipedia data.     

MATLAB在高等数学中的应用

MATLAB是矩阵实验室的简称,在图形处理方面表现突出。将MATLAB应用于高等数学的教学中,特别是在泰勒公式、空间解析几何以及极限理论的教学中,充分使用MATLAB的制图功能,能使抽象、枯燥的高等数学课程学习变得直观、明了和有趣。     

基于变分自编码器的异构缺陷预测特征表示方法

跨项目软件缺陷预测技术可以利用现有的已标注缺陷数据集对新的无标记项目进行预测,但需要两者之间具有相同的度量集合,难以用于实际开发.异构缺陷预测技术可以在具有异构度量集合的项目间进行缺陷预测,该技术引起了大量研究人员的关注.现有的异构缺陷预测技术利用朴素的或者传统机器学习方法为源项目和目标项目学习特征表示,所学习到的特征表示能力很弱且缺陷预测性能很差.鉴于深度神经网络强大的特征抽取和表示能力,本文基于变分自编码器技术提出了一种面向异构缺陷预测的特征表示方法.该模型结合了变分自编码器和最大均值差异距离,能有效地学习源项目和目标项目的共性特征表示,基于该特征表示可以训练出有效的缺陷预测模型.在多组缺陷数据集上通过与传统跨项目缺陷预测方法及异构缺陷预测方法实验对比验证了所提方法的有效性.     

基于元路径与层次注意力的时序异质信息网络表示学习方法

异质信息网络表示学习在节点分类、链接预测、个性化推荐等多个领域上被广泛应用.现有的异质信息网络表示学习方法大多集中在静态网络,忽略网络中时间属性对节点表示的影响.为了解决该问题,文中提出基于元路径和层次注意力的时序异质信息网络表示学习方法.利用元路径捕获异质信息网络中的结构和语义信息.通过时间衰减注意力层,捕获不同元路径实例在特定时间对目标节点的影响.通过元路径级别注意力,融合不同元路径下的节点表示,得到最终表示.在DBLP、IMDB数据集上的实验表明,文中方法在节点分类和节点聚类任务上均可达到较优效果.     

Spatial representation of symbolic sequences through iterativefunction systems

Jeffrey proposed (1990) a graphic representation of DNA sequences using Barnsley's iterative function systems. In spite of further developments in this direction, the proposed graphic representation of DNA sequences has been lacking a rigorous connection between its spatial scaling characteristics and the statistical characteristics of the DNA sequences themselves. We 1) generalize Jeffrey's graphic representation to accommodate (possibly infinite) sequences over an arbitrary finite number of symbols; 2) establish a direct correspondence between the statistical characterization of symbolic sequences via Renyi entropy spectra (1959) and the multifractal characteristics (Renyi generalized dimensions) of the sequences' spatial representations; 3) show that for general symbolic dynamical systems, the multifractal f_H-spectra in the sequence space coincide with the f_H -spectra on spatial sequence representations     

基于生成对抗模型的异质信息网络语义表征方法研究

近些年,网络表示学习问题吸引了大量研究者的关注,而异构信息网络由于其丰富的结构语义信息及其广阔的应用领域,更是成为了网络表示学习领域的重中之重。目前面向异构信息网络的表示学习模型主要可以分为基于生成式模型的表示学习方法和基于判别式模型的表示学习方法,但是很少有工作同时结合两种模型进行表示学习的优化。该文提出了结合生成式模型和判别式模型的异构信息网络表示学习模型HINGAN,主要是将对抗生成思想融入异构信息网络表示学习过程中,达到优化网络表示结果的目的。该模型首先在元路径的引导下构建带权信息网络图,然后在带权图上计算更新构造的生成器和判别器参数,通过生成对抗的博弈思想来获取最大收益。在AMiner和DBLP两个真实学术图谱数据集上的实验结果表明,HINGAN在多标签分类、链路预测以及可视化方面都能比现在主流的网络表示方法取得更优的效果,并且HINGAN可以应用于大规模的异构网络数据的表示和计算。除此之外,该文还总结了已有研究成果并对未来研究可能面临的挑战进行了展望。     

Effects of self-generated graphic organizers on learning depend on in-task guidance

We investigated the effects of readymade versus self-generated graphic organizers (GOs) on learning, comparing the performances of undergraduates (N = 81) tasked with learning a multimedia document. This document was either presented on its own (control group), with a readymade GO, or with a blank GO that students had to fill in either before or during the learning of the document. In line with previous research, adding a readymade GO increased students' memorization and transfer scores, compared with controls. By displaying the main ideas in the text and their hierarchical relations, GOs act as visual aids to learning. Results showed that self-generating a GO was no more beneficial than viewing a readymade GO when students were placed in a dual-task situation (generation + learning). However, when the students' information processing was guided by sequencing these tasks (generation then learning), they outperformed the control and readymade groups on memorization and comprehension.     

Visual tools for business model innovation: Recommendations from a cognitive perspective

Managers involved in business model innovation (BMI) encounter a series of cognitive challenges. Although cognition research provides important insights into how visual tools can reduce cognitive challenges, it does not address the effective design of business model tools. To advance our understanding in this area, this research applies a cognition‐centered perspective to analyze different visual business model representations. It builds upon a systematic literature review that identifies a selection of 45 visual representations from the academic literature and a sample of 50 visualizations from outside the academic realm. A content analysis of the sample reveals that all visual business model representations can be classified into three complementary categories, leading to a novel framework for distinguishing business model understandings. After assigning each visual representation to the framework, we use findings from the cognition literature to derive recommendations on how to select suitable graphic forms for different phases of the BMI process. Thus, this research contributes to the broader understanding of how visual tools can support business model innovation at a cognitive level.     

A unified interprocedural program representation for a maintenanceenvironment

Unified interprocedural graph (UIG) that extracts the important features of existing program representations and adds new information to provide an integrated representation for maintenance tasks is presented. Algorithms that were developed for previous representations are adapted to use the UIG by identifying the subset of nodes and edges in the UIG required for that computation. Newly developed algorithms can use the UIG since it contains data flow, control flow, data dependence, and control dependence information. The main benefits of this approach are the reduction in storage space since individual representations are not kept, the savings in maintenance time of a single representation over the individual representations, and the convenience of accessing a single program representation without increase in access time. A single program representation also assists in program understanding since relationships among program elements are incorporated into one graph     

Sharing in teams of heterogeneous,collaborative learning agents

This paper is focused on the effects of sharing knowledge and collaboration of multiple heterogeneous, intelligent agents (hardware or software) which work together to learn a task. As each agent employs a different machine learning technique, the system consists of multiple knowledge sources and their respective heterogeneous knowledge representations. Collaboration between agents involves sharing knowledge to both speed up team learning, as well as refine the team's overall performance and group behavior. Experiments have been performed that vary the team composition in terms of machine learning algorithms, learning strategies employed by the agents, and sharing frequency for a predator‐prey cooperative pursuit task. For lifelong learning, heterogeneous learning teams were more successful than homogeneous learning counterparts. Interestingly, sharing increased the learning rate, but sharing with higher frequency showed diminishing results. Lastly, knowledge conflicts are reduced over time the more sharing takes place. These results support further investigation of the merits of heterogeneous learning. © 2008 Wiley Periodicals, Inc.     

Superposition laws for solutions of differential matrix Riccati equations arising in control theory

Representation formulas are given for the general solution of theN times Nmatrix Riccati equationdot{W} = A + WB + CW + WDWusingnknown solutions, withn = 1, ..., 5(n-representations). The 5- representation is a superposition formula, in that it expresses the general solution explicitly as a function of five particular solutions and N²arbitrary constants (N geq 2), using no further information. The representation formulas can be used in numerical calculations. The 4- and 5- representations are specially useful when a solutionW(t)has a singularity for some finitet = t_{0}. They also clarify the properties of the solution space: the matrix elements ofW(t)are meromorphic functions ofthaving simple poles as the only possible singularities. The relation between the representation formulas and previously known results is discussed.     

基于门控图注意力网络的归纳式文本分类

为了有效地整合文本中的复杂特征和提取不同的上下文信息,提出了基于门控图注意力网络的归纳式文本分类方法 (TextIGAT).该方法首先为语料库中的每个文档进行单独构图,并将其中所有的单词作为图中的节点,以此保留完整的文本序列.文本图中设计单向连接的文档节点,使词节点能与全局信息交互,并合并不同的上下文关系连接词节点,从而在单个文本图中引入更多的文本信息.然后,方法基于图注意力网络(GAT)和门控循环单元(GRU)来更新词节点的表示,并根据图中保留的文本序列应用双向门控循环单元(Bi-GRU)来增强节点的顺序表示. TextIGAT能灵活地整合来自文本本身的信息,因此能对包含新词和关系的文本进行归纳式学习.在4个基准数据集(MR、Ohsumed、R8、R52)上的大量实验和详细分析表明了所提出的方法在文本分类任务上的有效性.     

Multimedia learning in social sciences: limitations of external graphical representations

In a series of six experimental studies, each consisting of three sub-studies, the central question was researched whether adding external graphical representations to printed or electronic learning materials improves retention and transfer scores. These studies research the degree of generalizability of Mayer’s cognitive theory of multimedia learning (CTML) to the knowledge domain of the social sciences. The research hypotheses build on the assumption that this knowledge domain differs in the way instructional designers are able to develop adequate depictive external graphical representations. Earlier CTML-research was mostly carried out in the field of the natural sciences, where graphical representations are depictive in nature and/or where representations can be developed from existing or acquired iconic sign systems. The results indicate that alternative guidelines might need to be considered when learners study learning materials with external graphical representations that reflect low levels of repleteness and do not build on an iconic sign system previously mastered or acquired by the learners. The research results reveal that studying this type of representation does not result in higher test performance and does not result in lower levels of mental load.