Learning and discovery: one system's search for mathematical knowledge

The Graph Theorist, GT, is a system that performs mathematical research in graph theory. From the definitions in its input knowledge base, GT constructs examples of mathematical concepts, conjectures and proves mathematical theorems about concepts, and discovers new concepts. Discovery is driven both by examples and by definitional form. The discovery processes construct a semantic net that links all of GT's concepts together.
Each definition is an algebraic expression whose semantic interpretation is a stylized algorithm to generate a class of graphs correctly and completely. From a knowledge base of these concept definitions, GT is able to conjecture and prove such theorems as "The set of acyclic, connected graphs is precisely the set of trees" and "There is no odd-regular graph on an odd number of vertices." GT explores new concepts either to develop an area of knowledge or to link a newly acquired concept into a pre-existing knowledge base. New concepts arise from the specialization of an existing concept, the generalization of an existing concept, and the merger of two or more existing concepts. From an initial knowledge base containing only the definition of "graph," GT discovers such concepts as acyclic graphs, connected graphs, and bipartite graphs.     

知识图谱综述——表示、构建、推理与知识超图理论

针对知识图谱(KG)在知识驱动的人工智能研究中发挥的强大支撑作用,分析并总结了现有知识图谱和知识超图技术.首先,从知识图谱的定义与发展历程出发,介绍了知识图谱的分类和架构;其次,对现有的知识表示与存储方式进行了阐述;然后,基于知识图谱的构建流程,分析了各类知识图谱构建技术的研究现状.特别是针对知识图谱中的知识推理这一重...     

图表示学习方法研究综述

针对图表示方法的相关解析任务进行了研究,从形式化定义出发,首先以不同核心技术作为分类标准将图表示学习方法划分为五大类,其包括基于降维解析、矩阵分解、随机游走、深度学习和其他表示学习方法。其次通过归纳与对比分析梳理各类技术发展脉络,进而深层次展现各类图表示方法的优劣。随后结合图表示学习的常用数据集、评估方法和应用领域的归纳分析,展开动态性、可扩展性、可解释性和可解析性的四维剖析。最后总结并展望了图表示学习的未来研究趋势与发展方向。     

基于图划分抽样算法的图表示学习

在基于神经网络的图表示算法中,当节点属性维度过高、图的规模过大时,从内存到显存的数据传输会成为训练性能的瓶颈。针对这类问题,该方法将图划分算法应用于图表示学习中,降低了内存访问的I/O开销。该方法根据图节点的度数,将图划分成若干个块,使用显存缓存池存储若干个特征矩阵块。每一轮训练,使用缓存池中的特征矩阵块,以此来减少内存到显存的数据拷贝。针对这一思想,该方法使用基于图划分的抽样算法,设计显存的缓存池来降低内存的访问,运用多级负采样算法,降低训练中负样本采样的时间复杂度。在多个数据集上,与现有方法对比发现,该方法的下游机器学习准确率与原算法基本一致,训练效率可以提高2~ 7倍。实验结果表明,基于图划分的图表示学习能高效训练模型,同时保证节点表示向量的测试效果。今后的课题可以使用严谨的理论证明,阐明图划分模型与原模型的理论误差。     

Knowledge acquisition and knowledge representation with class: the object-oriented paradigm

Knowledge acquisition and knowledge representation are the fundamental building blocks of knowledge-based systems (KBSs). How to efficiently elicit knowledge from experts and transform this elicited knowledge into a machine usable format is a significant and time consuming problem for KBS developers. Object-orientation provides several solutions to persistent knowledge acquisition and knowledge representation problems including transportability, knowledge reuse, and knowledge growth. An automated graphical knowledge acquisition tool is presented, based upon object-oriented principles. The object-oriented graphical interface provides a modeling platform that is easily understood by experts and knowledge engineers. The object-oriented base for the automated KA tool provides a representation independent methodology that can easily be mapped into any other object-oriented expert system or other object-oriented intelligent tools.     

Knowledge discovery techniques for predicting country investment risk

This paper presents the insights gained from applying knowledge discovery in databases (KDD) processes for the purpose of developing intelligent models, used to classify a country's investing risk based on a variety of factors. Inferential data mining techniques, like C5.0, as well as intelligent learning techniques, like neural networks, were applied to a dataset of 52 countries. The dataset included 27 variables (economic, stock market performance/risk and regulatory efficiencies) on 52 countries, whose investing risk category was assessed in a Wall Street Journal survey of international experts. The results of applying KDD techniques to the dataset are promising, and successfully classified most countries as compared to the experts' classifications. Implementation details, results, and future plans are also presented.     

引入知识表示的图卷积网络谣言检测方法

在谣言检测的问题上,现有的研究方法无法有效地表达谣言在社交网络传播的异构图结构特征,并且没有引入外部知识作为内容核实的手段。因此,提出了引入知识表示的图卷积网络谣言检测方法,其中知识图谱作为额外先验知识来帮助核实内容真实性。采用预训练好的词嵌入模型和知识图谱嵌入模型获取文本表示后,融合图卷积网络的同时,能够在谣言传播的拓扑图中更好地进行特征提取以提升谣言检测的精确率。实验结果表明,该模型能够更好地对社交网络中的谣言进行检测。与基准模型的对比中,在Weibo数据集上的精确率达到96.1%,在Twitter15和Twitter16数据集上的F1值分别提升了3.1%和3.3%。消融实验也表明了该方法对谣言检测皆有明显提升效果,同时验证了模型的有效性和先进性。     

Knowledge discovery standards

As knowledge discovery (KD) matures and enters the mainstream, there is an onus on the technology developers to provide the technology in a deployable, embeddable form. This transition from a stand-alone technology, in the control of the knowledgeable few, to a widely accessible and usable technology will require the development of standards. These standards need to be designed to address various aspects of KD ranging from the actual process of applying the technology in a business environment, so as to make the process more transparent and repeatable, through to the representation of knowledge generated and the support for application developers. The large variety of data and model formats that researchers and practitioners have to deal with and the lack of procedural support in KD have prompted a number of standardization efforts in recent years, led by industry and supported by the KD community at large. This paper provides an overview of the most prominent of these standards and highlights how they relate to each other using some example applications of these standards.     

基于本体和二部图的网格服务发现算法研究

为了提高网格服务发现的查全率、查准率和效率,论文设计了一个基于本体和二部图的网格服务发现算法OGSDA-BG。该算法把请求服务和发布服务的属性集分别作为二部图顶点集,所有匹配属性之间的连线为边,边权是属性匹配度,把问题转换为二部图的最优完全匹配。实验结果表明该算法的查全率和查准率较以前的算法提高了10%~50%,尽管服务发现的效率降低10%左右,但是在可接受范围之内。     

Knowledge graph embeddings for dealing with concept drift in machine learning

Data stream learning has been largely studied for extracting knowledge structures from continuous and rapid data records. As data is evolving on a temporal basis, its underlying knowledge is subject to many challenges. Concept drift,¹ as one of core challenge from the stream learning community, is described as changes of statistical properties of the data over time, causing most of machine learning models to be less accurate as changes over time are in unforeseen ways. This is particularly problematic as the evolution of data could derive to dramatic change in knowledge. We address this problem by studying the semantic representation of data streams in the Semantic Web, i.e., ontology streams. Such streams are ordered sequences of data annotated with ontological vocabulary. In particular we exploit three levels of knowledge encoded in ontology streams to deal with concept drifts: i) existence of novel knowledge gained from stream dynamics, ii) significance of knowledge change and evolution, and iii) (in)consistency of knowledge evolution. Such knowledge is encoded as knowledge graph embeddings through a combination of novel representations: entailment vectors, entailment weights, and a consistency vector. We illustrate our approach on classification tasks of supervised learning. Key contributions of the study include: (i) an effective knowledge graph embedding approach for stream ontologies, and (ii) a generic consistent prediction framework with integrated knowledge graph embeddings for dealing with concept drifts. The experiments have shown that our approach provides accurate predictions towards air quality in Beijing and bus delay in Dublin with real world ontology streams.     

知识图谱构建技术:分类、调查和未来方向

知识图谱的概念由谷歌于2012年提出,随后逐渐成为人工智能领域的一个研究热点,已在信息搜索、自动问答、决策分析等应用中发挥作用。虽然知识图谱在各领域展现出了巨大的潜力,但不难发现目前缺乏成熟的知识图谱构建平台,需要对知识图谱的构建体系进行研究,以满足不同的行业应用需求。文中以知识图谱构建为主线,首先介绍目前主流的通用知识图谱和领域知识图谱,描述两者在构建过程中的区别;然后,分类讨论图谱构建过程中存在的问题和挑战,并针对这些问题和挑战,分类描述目前图谱构建过程中的知识抽取、知识表示、知识融合、知识推理、知识存储5个层面的解决方法和策略;最后,展望未来可能的研究方向。     

Knowledge discovery in data sets with graded attributes

We present a knowledge discovery method for graded attributes that is based on an interactive determination of implications (if-then-rules) holding between the attributes of a given data-set. The corresponding algorithm queries the user in an efficient way about implications between the attributes. The result of the process is a representative set of examples for the entire theory and a set of implications from which all implications that hold between the attributes can be deduced. In many instances, the exploration process may be shortened by the usage of the user’s background knowledge. That is, a set of of implications the user knows beforehand. The method was successfully applied in different real-life applications for discrete data. In this paper, we show that attribute exploration with background information can be generalized for graded attributes.     

一种新的关联规则Vis-Meta图知识表示方法

针对传统关联规则表示方法无法展示领域知识,缺少对规则信息一对一、一对多、多对一、多对多的多模式表示,忽略知识发现结果的共享等问题,提出了一种新的基于Vis-Meta图的多模式关联规则知识表示方法.首先给出了Vis-Meta图的相关定义与关联规则的Vis-Meta图表示方法,接着定义了关联规则Vis-Meta图知识表示中的概念关系,并在此基础上给出了关联规则概念关系知识表示算法、关联规则实例对比算法和关联规则知识表示优化算法.最后,以某省全员人口数据为基础,对关联规则信息进行可视化分析.实验结果表明,所提出的知识表示算法具有良好的展示效果与知识共享能力.     

Neural networks and graph theory

The relationships between artificial neural networks and graph theory are considered in detail. The applications of artificial neural networks to many difficult problems of graph theory, especially NP-complete problems, and the applications of graph theory to artificial neural networks are discussed. For example graph theory is used to study the pattern classification problem on the discrete type feedforward neural networks, and the stability analysis of feedback artificial neural networks etc.     

Applications of graph theory in computer systems

Many problem situations in computer systems can be analyzed using models based on directed graphs. The vertices of the graph represent states of the system and the directed arcs represent the transitions between these states. This paper is in two parts. The first introduces the concepts of directed graphs and their representations in computers and presents some basic problems and algorithms. The second part examines the application of graph theory to various areas of computer systems.     

四元数关系旋转的知识图谱补全模型

知识图谱是真实世界三元组的结构化表示,通常三元组被表示成头实体、关系、尾实体的形式.针对知识图谱中广泛存在的数据稀疏问题,提出了一种将四元数作为关系旋转的知识图谱补全方法.文中使用极具表现力的超复数表示对实体和关系进行建模,以进行链接预测.这种超复数嵌入用于表示实体,关系则被建模为四元数空间中的旋转.具体来说,将每个关...     

Knowledge discovery in corporate events by neural network rule extraction

Over the last two decades, artificial neural networks (ANN) have been applied to solve a variety of problems such as pattern classification and function approximation. In many applications, it is desirable to extract knowledge from trained neural networks for the users to gain a better understanding of the network’s solution. In this paper, we use a neural network rule extraction method to extract knowledge from 2222 dividend initiation and resumption events. We find that the positive relation between the short-term price reaction and the ratio of annualized dividend amount to stock price is primarily limited to 96 small firms with high dividend ratios. The results suggest that the degree of short-term stock price underreaction to dividend events may not be as dramatic as previously believed. The results also show that the relations between the stock price response and firm size is different across different types of firms. Thus, drawing the conclusions from the whole dividend event data may leave some important information unexamined. This study shows that neural network rule extraction method can reveal more knowledge from the data.     

The evolutionary perspective of knowledge creation – A mathematical representation

Since knowledge and its strategic applications have now become a major source of competitive advantage for organizations, the lifeblood of organizations is to possess the capabilities for acquiring, creating, sharing, diffusing, utilizing, and storing knowledge among organizational members. Among the capabilities, knowledge creation is the major resource of organizational innovation and it, therefore, plays a more crucial role in developing a sustained competitive advantage for organizations, especially in a dynamic environment. Although numerous papers have studied the issue of knowledge creation, those papers focused on conceptualizing the process of knowledge creation and did not investigate the concrete problem of knowledge creation. In the current paper, a mathematical representation of knowledge creation based on an evolutionary perspective is used to understand: (1) if knowledge can be created; (2) how knowledge is created; (3) under what conditions knowledge can be created; (4) the partial effects of variables which affect the process of knowledge creation.     

一种基于新词发现的Web文本表示方法

Web文本表示是Web文本特征提取和分类的前提,最常用的文本表示是向量空间模型（VSM）,其中向量一般是基于词的特征项。由于向量空间模型本身没有考虑文本上下文间的潜在概念结构（如词汇间的共现关系）,而Web文本是一种半结构化文本,同时经常有新词出现,因此在VSM基础上提出了一种基于新词发现的Web文本表示方法：首先进行预处理将网页转化为文本;然后进行文本分词;接着通过二元互信息进行新词发现,同时把新词加入字典重新分词;最后用词和新词共同来表示Web文本。实验结果表明,该方法可以帮助识别未登录词并扩充现有字典,能够增强Web文本表示能力,改善Web文本的特征项质量,提高Web文本分类效果。