Future Event Prediction Based on Temporal Knowledge Graph Embedding

Accurate prediction of future events brings great benefits and reduces losses for society in many domains,such as civil unrest,pandemics,and crimes.Knowledge graph is a general language for describing and modeling complex systems.Different types of events continually occur,which are often related to historical and concurrent events.In this paper,we formalize the future event prediction as a temporal knowledge graph reasoning problem.Most existing studies either conduct reasoning on static knowledge graphs or assume knowledges graphs of all timestamps are available during the training process.As a result,they cannot effectively reason over temporal knowledge graphs and predict events happening in the future.To address this problem,some recent works learn to infer future events based on historical eventbased temporal knowledge graphs.However,these methods do not comprehensively consider the latent patterns and influences behind historical events and concurrent events simultaneously.This paper proposes a new graph representation learning model,namely Recurrent Event Graph ATtention Network(RE-GAT),based on a novel historical and concurrent events attention-aware mechanism by modeling the event knowledge graph sequence recurrently.More specifically,our RE-GAT uses an attention-based historical events embedding module to encode past events,and employs an attention-based concurrent events embedding module to model the associations of events at the same timestamp.A translation-based decoder module and a learning objective are developed to optimize the embeddings of entities and relations.We evaluate our proposed method on four benchmark datasets.Extensive experimental results demonstrate the superiority of our RE-GAT model comparing to various base-lines,which proves that our method can more accurately predict what events are going to happen.

News event prediction by trigger evolution graph and event segment

Event prediction aims to predict the most possible following event given a chain of closely related context events.Previous methods based on event pairs or the entire event chain may ignore much structural and semantic information.Current datasets for event prediction,naturally,can be used for supervised learning.Event chains are either from document-level procedural action flow,or from news sequences under the same column.This paper leverages graph structure knowledge of event triggers and event segment information for event prediction with general news corpus,and adopts the standard multiple choice narrative cloze task evaluation.The topic model is utilized to extract event chains from the news corpus to deal with training data bottleneck.Based on trigger-guided structural relations in the event chains,we construct trigger evolution graph,and trigger representations are learned through graph convolutional neural network and the novel neighbor selection strategy.Then there are features of two levels for each event,namely,text level semantic feature and trigger level structural feature.We design the attention mechanism to learn the features of event segments derived in term of event major subjects,and integrate relevance between event segments and the candidate event.The most possible next event is picked by the relevance.Experimental results on the real-world news corpus verify the effectiveness of the proposed model.

Application Research on Two-Layer Threat Prediction Model Based on Event Graph

Advanced Persistent Threat(APT)is now the most common network assault.However,the existing threat analysis models cannot simultaneously predict the macro-development trend and micro-propagation path of APT attacks.They cannot provide rapid and accurate early warning and decision responses to the present system state because they are inadequate at deducing the risk evolution rules of network threats.To address the above problems,firstly,this paper constructs the multi-source threat element analysis ontology(MTEAO)by integrating multi-source network security knowledge bases.Subsequently,based on MTEAO,we propose a two-layer threat prediction model(TL-TPM)that combines the knowledge graph and the event graph.The macro-layer of TL-TPM is based on the knowledge graph to derive the propagation path of threats among devices and to correlate threat elements for threat warning and decision-making;The micro-layer ingeniously maps the attack graph onto the event graph and derives the evolution path of attack techniques based on the event graph to improve the explainability of the evolution of threat events.The experiment’s results demonstrate that TL-TPM can completely depict the threat development trend,and the early warning results are more precise and scientific,offering knowledge and guidance for active defense.

Visibility graph approach to extreme event series

An extreme event may lead to serious disaster to a complex system.In an extreme event series there exist generally non-trivial patterns covering different time scales.Investigations on extreme events are currently based upon statistics,where the patterns are merged into averages.In this paper from extreme event series we constructed extreme value series and extreme interval series.And the visibility graph is then adopted to display the patterns formed by the increases/decreases of extreme value or interval faster/slower than the linear ones.For the fractional Brownian motions,the properties for the constructed networks are the persistence,threshold,and event-type-independent,e.g.,the degree distributions decay exponentially with almost identical speeds,the nodes cluster into modular structures with large and similar modularity degrees,and each specific network has a perfect hierarchical structure.For the volatilities of four stock markets(NSDQ,SZI,FTSE100,and HSI),the properties for the former three's networks are threshold-and market-independent.Comparing with the factional Brownian motions,their degree distributions decay exponentially but with slower speeds,their modularity behaviors are significant but with smaller modularity degrees.The fourth market behaves similar qualitatively but different quantitatively with the three markets.Interestingly,all the transition frequency networks share an identical backbone composed of nine edges and the linked graphlets.The universal behaviors give us a framework to describe extreme events from the viewpoint of network.

Combing Type-Aware Attention and Graph Convolutional Networks for Event Detection

Event detection(ED)is aimed at detecting event occurrences and categorizing them.This task has been previously solved via recognition and classification of event triggers(ETs),which are defined as the phrase or word most clearly expressing event occurrence.Thus,current approaches require both annotated triggers as well as event types in training data.Nevertheless,triggers are non-essential in ED,and it is time-wasting for annotators to identify the“most clearly”word from a sentence,particularly in longer sentences.To decrease manual effort,we evaluate event detectionwithout triggers.We propose a novel framework that combines Type-aware Attention and Graph Convolutional Networks(TA-GCN)for event detection.Specifically,the task is identified as a multi-label classification problem.We first encode the input sentence using a novel type-aware neural network with attention mechanisms.Then,a Graph Convolutional Networks(GCN)-based multilabel classification model is exploited for event detection.Experimental results demonstrate the effectiveness.

Modeling of unsupervised knowledge graph of events based on mutual information among neighbor domains and sparse representation

Text event mining,as an indispensable method of text mining processing,has attracted the extensive attention of researchers.A modeling method for knowledge graph of events based on mutual information among neighbor domains and sparse representation is proposed in this paper,i.e.UKGE-MS.Specifically,UKGE-MS can improve the existing text mining technology's ability of understanding and discovering high-dimensional unmarked information,and solves the problems of traditional unsupervised feature selection methods,which only focus on selecting features from a global perspective and ignoring the impact of local connection of samples.Firstly,considering the influence of local information of samples in feature correlation evaluation,a feature clustering algorithm based on average neighborhood mutual information is proposed,and the feature clusters with certain event correlation are obtained;Secondly,an unsupervised feature selection method based on the high-order correlation of multi-dimensional statistical data is designed by combining the dimension reduction advantage of local linear embedding algorithm and the feature selection ability of sparse representation,so as to enhance the generalization ability of the selected feature items.Finally,the events knowledge graph is constructed by means of sparse representation and l1 norm.Extensive experiments are carried out on five real datasets and synthetic datasets,and the UKGE-MS are compared with five corresponding algorithms.The experimental results show that UKGE-MS is better than the traditional method in event clustering and feature selection,and has some advantages over other methods in text event recognition and discovery.

Development of an Improved GUI Automation Test System Based on Event-Flow Graph

A more automated graphic user interface (GUI) test model, which is based on the event-flow graph, is proposed. In the model, a user interface automation API tool is first used to carry out reverse engineering for a GUI test sample so as to obtain the event-flow graph. Then two approaches are adopted to create GUI test sample cases. That is to say, an improved ant colony optimization (ACO) algorithm is employed to establish a sequence of testing cases in the course of the daily smoke test. The sequence goes through all object event points in the event-flow graph. On the other hand, the spanning tree obtained by deep breadth-first search (BFS) approach is utilized to obtain the testing cases from goal point to outset point in the course of the deep regression test. Finally, these cases are applied to test the new GUI. Moreover, according to the above-mentioned model, a corresponding prototype system based on Microsoft UI automation framework is developed, thus giving a more effective way to improve the GUI automation test in Windows OS.

Cyclic Reconfigurable Flow Shop under Different Configurations Modeling and Optimization Based on Timed Event Graph

Based on the idea that modules are independent of machines, different combinations of modules and machines result in different configurations and the system performances differ under different configurations, a kind of cyclic reconfigurable flow shops are proposed for the new manufacturing paradigm-reconfigurable manufacturing system. The cyclic reconfigurable flow shop is modeled as a timed event graph. The optimal configuration is defined as the one under which the cyclic reconfigurable flow shop functions with the minimum cycle time and the minimum number of pallets. The optimal configuration, the minimum cycle time and the minimum number of pallets can be obtained in two steps.

以评促教:基于事理图谱的计算思维水平评价方法

计算思维是人工智能时代教育发展的核心目标,而监测计算思维状况、评价计算思维水平并据此提出计算思维培养的教学策略是当前“教-学-评”一体化教学改革下实现计算思维培养的重要路径。然而,由于计算思维的复杂性、抽象性和过程性,教师往往难以全面掌握和准确评价学生计算思维应用过程,更难以个性化了解学生的计算思维发展状况并据此提出教学干预策略。为此,从测评的视角出发,基于计算思维的本质内涵,结合新兴的事理图谱概念,构建基于事理图谱的计算思维评价方式。首先,设计面向计算思维的事件编码框架,利用iBCM算法实现事件关系抽取,生成事理图谱。然后,利用随机森林根据事理图谱对计算思维水平进行分类,并采用基于规则的事理图谱优化方法,完善iBCM提取出的特征模式,生成可视化的事理图谱辅助教学策略的调整与个性化的指导。基于此方法开展的实证研究表明,设计的模型在各项指标上都具有良好的性能效果。其结合了事理图谱所具备的事件逻辑深层次分析与推理的优势,为理解学习者的思维过程、抽取思维特征、探究思维模式提供了支撑和条件,助力教师培养学习者的计算思维提供路径指导。

融合情感语义与句法结构的中文开放域事理图谱构建研究

为解决大规模开放域事理图谱构建过程中缺少标注数据以及事件类型未知导致的限定域事理图谱构建方法难以迁移的问题,利用规则匹配方法高效识别开放域文本中包含的多种事件逻辑关系,融合情感语义与句法结构信息分析提高事件抽取准确性,以更好完成事理图谱的构建任务。首先,总结并扩展因果、顺承、条件、转折等多种逻辑关系抽取模板,并基于规则模板、依存句法信息筛选逻辑关系事件句;其次,创新性地引入情感语义分析方法,在句法结构信息的基础上,通过捕获事件及事件间关系的情感语义精准识别事件类型,进而抽取事件论元;再次,计算语义相似度,进行事件融合,构建<前序事件,事件逻辑关系,后序事件>三元组,得到事件事理图谱,并进一步进行事件泛化以构建抽象事理图谱;最后,以事件发展较完整的“2022年猴痘事件”为数据源,通过实证分析证明开放域事理图谱构建方法可以实现不同类型事件的识别、事件间逻辑关系的揭露,其有效性、可行性得到验证。研究不仅弥补了现有事理图谱构建理论的不足,也为决策支持、事件发展预测等提供有力的数据支持。

基于知识图谱的冬奥赛事气象服务文本生成方法研究

气象服务文本是为赛事顺利举行及赛事期间的各项活动提供必要的气象保障,是组委会、裁判、各代表队的工作人员获取气象信息开展相关工作的载体。现有气象文本生产需要人工编写审核,效率不高。相比之下,全自动文本生成更加依赖于模板和固定的形式。针对以上问题,结合自然语言处理技术提出基于知识图谱的冬奥赛事气象服务文本生成方法。重点从历史赛事气象服务文本中进行内容分析和特征提取,利用气象数据和历史赛事信息构建高山滑雪赛事知识图谱。该方法根据实时气象数据和文稿模板生成天气描述文本,然后基于知识图谱查询推理技术得到赛事影响结果并生成相应文本。实验结果表明:气象服务文本的自动生成结果具有较好的准确性和可读性,有助于冬奥赛事的顺利推进,该文本生成方法面向特定领域也具有较好的应用前景。

基于多序列隐关系的时序事件预测

时序事件预测是指基于历史事件预测下一个事件,事件包括时间和类型两个属性。当前主要工作集中在单方面(事件时间或事件类型)的预测,但这无法回答“何时发生何事”这类更精细的问题。此类问题的挑战主要是事件类型非常多样,而行为往往高度稀疏,给预测带来极大困难;需要预测的事件时间和事件类型分属两个域,如何把这两个域的信息加以融合并形成互补也是一个挑战。针对上述挑战,从融合多序列隐信息的角度探索了一种解决方法。基于某些事件序列之间具有模式相似性这一观察,提出建模事件序列的隐关系图,利用邻居序列的信息解决行为稀疏性的问题;通过合理设计神经网络模块,将事件的时间域和类型域的信息映射到共同的抽象空间,解决事件时间和事件类型信息的融合建模问题。通过在多个真实数据集上进行了大量实验,实验结果印证了多序列深度时序模型优于现有的一系列基准模型。

基于多粒度阅读器和图注意力网络的文档级事件抽取

文档级事件抽取面临论元分散和多事件两大挑战,已有工作大多采用逐句抽取候选论元的方式,难以建模跨句的上下文信息。为此,提出了一种基于多粒度阅读器和图注意网络的文档级事件抽取模型,采用多粒度阅读器实现多层次语义编码,通过图注意力网络捕获实体对之间的局部和全局关系,构建基于实体对相似度的剪枝完全图作为伪触发器,全面捕捉文档中的事件和论元。在公共数据集ChFinAnn和DuEE-Fin上进行了实验,结果表明提出的方法改善了论元分散问题,提升了模型事件抽取性能。

基于事件演化图的多标记事件预测模型

多标记事件预测是指预测多个相关联的事件是否会在未来发生,相比传统单标记事件预测,需要同时预测多个目标事件。现有的事件预测研究忽略各领域存在的多标记事件情境,且对多标记事件预测研究较少。提出一种基于事件演化图的多标记事件预测模型(MLEP),以实现基于事件演化图(EEG)的多标记事件预测研究模式。首先基于事件链构建事件演化图;然后对多标记事件预测问题进行问题转换,将多标记问题转化为单标记问题,利用事件表示学习方法获取所有事件的向量表示,对多标记事件进行编码;最后采用门控图神经网络(GGNN)框架构建多标记事件预测模型,根据相似度匹配出最优的后续事件,实现多标记事件的预测。在真实数据集上的实验结果表明,MLEP模型可以有效地预测出多标记事件,预测准确率达到了65.58%,性能优于大多现有的基准模型,提升幅度达到了4.94%以上。通过消融实验也证明了更好的事件表示学习方法对事件具有较好的表示效果,提升多标记事件预测的性能。

分布式事件触发下的多智能体系统二分一致性

研究了基于分布式事件触发的多智能体控制系统中的二分一致性问题,设计了一种合作和竞争并存拓扑结构的一致性控制协议,使多智能体系统收敛于二模相同且符号不同的状态.在图论与矩阵论的基础上,将多智能体系统的二分一致性问题转化为闭环误差系统的稳定性问题,通过建立一个合适的Lyapunov函数,验证了闭环误差系统渐进稳定,且针对每个智能体提出了事件触发条件,以减少控制更新频率与通信资源的损耗,同时证明每个智能体相邻触发时间间隔都有严格的正下界,即避免了Zeno现象,最后利用仿真实例验证了所提方法的可行性和有效性.

多源知识图谱事件知识融合方法研究

以事件为中心的动态知识对事件预测等应用至关重要,但现有知识图谱主要关注以实体为中心的静态知识,难以满足需求。本文提出一种融合多源知识的高质量事件知识图谱构造方法,首先定义全局事件模式,利用标签类别从源知识图谱中提取事件知识并构造临时事件知识图,提出关系扩充规则对临时事件知识图进行扩充,改进实体对齐Attce模型,基于TransD模型对多个临时事件知识图进行联合嵌入学习,以提高实体对齐和冲突发现的效率;利用事件描述完整度计算源知识图谱可信度,发生冲突时作为判别标准进行处理。经过在真实数据集上的实验,验证了该方法的准确性和有效性。

面向信息公开的突发公共卫生事件知识表示模型构建

近年来,突发公共卫生事件频发给政府应急管理带来了极大挑战。政府及时公开事件信息有助于消除公众恐慌,对于疫情防控和社会经济发展至关重要。然而,当前突发公共卫生事件信息以碎片化、不连续、不完整方式散落在不同位置,未能较好地集成融合。如何对海量、多样、变化的信息进行描述与组织是政府应急管理的关键。因此,本文面向信息公开,融合知识图谱和事理图谱,构建突发公共卫生事件知识表示模型,并对其中的核心概念及关系进行表示,提出了一种既能刻画疫情时空演化状态,又能展示疫情信息的突发公共卫生事件信息组织方法。研究结果显示,本文构建的知识表示模型具有较好的垂直性质,能够展示较为丰富的概念关系和属性特征,大部分类能够得到实例填充。本文提出的知识表示模型为突发事件知识库构建提供新思路,扩充了突发事件信息组织的方法体系,有助于存量信息公开与价值释放,进一步满足公众的信息需求,提升应急管理效果。

网络服务异常事件告警因果图构造方法

网络服务系统中,异常事件的发生经常导致系统中产生大量告警事件,形成告警风暴。运维人员需要花费大量的时间和精力从这些告警数据中寻找关键信息、确定异常事件的根源。为了减少运维人员所需处理的告警数量,智能化、自动化地提取告警风暴中的根源告警,基于网络服务告警的传播模式分析,提出了一种告警因果图构造方法,并将其应用于提取异常事件发生时的告警风暴关键信息。实验使用运营商现网管理系统的真实数据集,通过告警风暴摘要提取实验,验证了告警因果图生成的效果,并进行了相关案例的物理意义分析。结果表明,使用告警因果图生成的方式进行告警风暴摘要提取,达到了96%的召回率,保留了绝大部分关键信息。同时,使用该方法对系统产生的告警进行压缩,对较难压缩的告警码的压缩率能够达到66.5%。

基于强化联邦GNN的个性化公共安全突发事件检测

近年来,将公共安全数据转换为图的形式,通过图神经网络(GNN)构造节点表示应用于下游任务的方法,充分利用了公共安全数据的实体与关联信息,取得了较好的效果.为了提高模型的有效性,需要大量的高质量数据,但是高质量的数据通常归属于政府、公司和组织,很难通过数据集中的方式使模型学习到有效的事件检测模型.由于各数据拥有方的关注主题与收集时间不同,数据之间存在Non-IID的问题.传统的假设一个全局模型可以适合所有客户端的方法难以解决此类问题.提出了基于强化联邦图神经网络的个性化公共安全突发事件检测方法PPSED,各客户端采用多方协作的方式训练个性化的模型来解决本地的突发事件检测任务.设计了联邦公共安全突发事件检测模型的本地训练与梯度量化模块,采用基于图采样的minibatch机制的GraphSage构造公共安全突发事件检测本地模型,以减小数据Non-IID的影响,采用梯度量化方法减小梯度通信的消耗.设计了基于随机图嵌入的客户端状态感知模块,在保护隐私的同时,更好地保留客户端模型有价值的梯度信息.设计了强化联邦图神经网络的个性化梯度聚合与量化策略,采用DDPG拟合个性化联邦学习梯度聚合加权策略,并根据权重决定是否对梯度进行量化,对模型的性能与通信压力进行平衡.通过在微博平台收集的公共安全数据集和3个公开的图数据集进行了大量的实验,实验结果表明了所提方法的有效性.

基于对比图学习的跨文档虚假信息检测

当前,网络上充斥着大量虚假信息,严重阻碍了社会各行业的正常运转,如何精准检测虚假信息成为了亟待解决的问题。现有研究主要从账户特征、文本内容和多模态3个角度开展工作,但大多忽视了虚假信息赖以传播的关键特征(即内容新奇性),仅是孤立地分析判别目标信息的真实性,未能把握舆论环境的特征。因此,提出了一种基于对比图学习的跨文档虚假信息检测方法(Contrastive Graph Learning,CAL),聚焦于内容新奇性,主要包含两个关键模块:对比学习模块和异构图模块。前者致力于扩大客观事实与虚假信息在向量空间中的表示差异性;后者包含实体、事件、事件集、句子和文档5种类型实体,尽可能向实体表示中注入舆论环境的语义特征。最后,在IED,TL17和Crisis这3个数据集上,在文档级和事件级这两个层次上开展了相关实验,CAL在所有测试中均取得了最优的结果,验证了所提方法的有效性。