Clustering incomplete relational data using the non-Euclidean relational fuzzy c-means algorithm

An approach for clustering on the basis of incomplete dissimilarity data is given. The data is first completed using simple triangle inequality-based approximation schemes and then clustered using the non-Euclidean relational fuzzy c-means algorithm. Results of numerical tests are included.     

基于聚类关联网络的群组行为识别

目前群组行为识别方法没有充分利用群组关联信息而导致群组识别精度无法有效提升,针对这个问题,提出了基于近邻传播算法（AP）的层次关联模块的深度神经网络模型,命名为聚类关联网络（CRN）。首先,利用卷积神经网络（CNN）提取场景特征,再利用区域特征聚集提取场景中的人物特征。然后,利用AP的层次关联网络模块提取群组关联信息。最后,利用长短期记忆网络（LSTM）融合个体特征序列与群组关联信息,并得到最终的群组识别结果。与多流卷积神经网络（MSCNN）方法相比,CRN方法在Volleyball数据集与Collective Activity数据集上的识别准确率分别提升了5.39与3.33个百分点。与置信度能量循环网络（CERN）方法相比,CRN方法在Volleyball数据集与Collective Activity数据集上的识别准确率分别提升了8.7与3.14个百分点。实验结果表明,CRN方法在群体行为识别任务中拥有更高的识别准确精度。     

基于聚类关联网络的群组行为识别

目前群组行为识别方法没有充分利用群组关联信息而导致群组识别精度无法有效提升，针对这个问题，提出了基于近邻传播算法（AP）的层次关联模块的深度神经网络模型，命名为聚类关联网络（CRN）。首先，利用卷积神经网络（CNN）提取场景特征，再利用区域特征聚集提取场景中的人物特征。然后，利用AP的层次关联网络模块提取群组关联信息。最后，利用长短期记忆网络（LSTM）融合个体特征序列与群组关联信息，并得到最终的群组识别结果。与多流卷积神经网络（MSCNN）方法相比，CRN方法在Volleyball数据集与Collective Activity数据集上的识别准确率分别提升了5.39与3.33个百分点。与置信度能量循环网络（CERN）方法相比，CRN方法在Volleyball数据集与Collective Activity数据集上的识别准确率分别提升了8.7与3.14个百分点。实验结果表明，CRN方法在群体行为识别任务中拥有更高的识别准确精度。     

Clustering and aggregation of relational data with applications to image database categorization

In this paper, we introduce a new algorithm for clustering and aggregating relational data (CARD). We assume that data is available in a relational form, where we only have information about the degrees to which pairs of objects in the data set are related. Moreover, we assume that the relational information is represented by multiple dissimilarity matrices. These matrices could have been generated using different sensors, features, or mappings. CARD is designed to aggregate pairwise distances from multiple relational matrices, partition the data into clusters, and learn a relevance weight for each matrix in each cluster simultaneously. The cluster dependent relevance weights offer two advantages. First, they guide the clustering process to partition the data set into more meaningful clusters. Second, they can be used in subsequent steps of a learning system to improve its learning behavior. The performance of the proposed algorithm is illustrated by using it to categorize a collection of 500 color images. We represent the pairwise image dissimilarities by six different relational matrices that encode color, texture, and structure information.     

Heterogeneous relational reasoning in knowledge graphs with reinforcement learning

Path-based relational reasoning over knowledge graphs has become increasingly popular due to a variety of downstream applications such as question answering in dialogue systems, fact prediction, and recommendation systems. In recent years, reinforcement learning (RL) based solutions for knowledge graphs have been demonstrated to be more interpretable and explainable than other deep learning models. However, the current solutions still struggle with performance issues due to incomplete state representations and large action spaces for the RL agent. We address these problems by developing HRRL (Heterogeneous Relational reasoning with Reinforcement Learning), a type-enhanced RL agent that utilizes the local heterogeneous neighborhood information for efficient path-based reasoning over knowledge graphs. HRRL improves the state representation using a graph neural network (GNN) for encoding the neighborhood information and utilizes entity type information for pruning the action space. Extensive experiments on real-world datasets show that HRRL outperforms state-of-the-art RL methods and discovers more novel paths during the training procedure, demonstrating the explorative power of our method.     

基于空间邻近的点目标聚类方法

空间聚类是空间数据挖掘和知识发现领域的主要研究方向之一,但点目标空间分布密度的不均匀、分布形状的多样化,以及"多桥"链接问题的存在,使得基于距离和密度的聚类算法不能高效且有效地识别聚集性高的点目标。提出了基于空间邻近的点目标聚类方法,通过Voronoi建模识别点目标间的空间邻近关系,并以Voronoi势力范围来定义相似度准则,最终构建树结构以实现点目标的聚集模式识别。实验将所提算法与K-means、具有噪声的基于密度的聚类(DBSCAN)算法进行比较分析,结果表明算法能够发现密度不均且任意形状分布的点目标集群,同时准确划分"桥"链接的簇,适用于空间点目标异质分布下的聚集模式识别。     

Fuzzy relational clustering based on comparing two proximity matrices with utilization of particle swarm optimization

The first stage of knowledge acquisition and reduction of complexity concerning a group of entities is to partition or divide the entities into groups or clusters based on their attributes or characteristics. Clustering algorithms normally require both a method of measuring proximity between patterns and prototypes and a method for aggregating patterns. However sometimes feature vectors or patterns may not be available for objects and only the proximities between the objects are known. Even if feature vectors are available some of the features may not be numeric and it may not be possible to find a satisfactory method of aggregating patterns for the purpose of determining prototypes. Clustering of objects however can be performed on the basis of data describing the objects in terms of feature vectors or on the basis of relational data. The relational data is in terms of proximities between objects. Clustering of objects on the basis of relational data rather than individual object data is called relational clustering. The premise of this paper is that the proximities between the membership vectors, which are obtained as the objective of clustering, should be proportional to the proximities between the objects. The values of the components of the membership vector corresponding to an object are the membership degrees of the object in the various clusters. The membership vector is just a type of feature vector. Based on this premise, this paper describes another fuzzy relational clustering method for finding a fuzzy membership matrix. The method involves solving a rather challenging optimization problem, since the objective function has many local minima. This makes the use of a global optimization method such as particle swarm optimization (PSO) attractive for determining the membership matrix for the clustering. To minimize computational effort, a Bayesian stopping criterion is used in combination with a multi-start strategy for the PSO. Other relational clustering methods generally find local optimum of their objective function.     

Inductive logic programming for relational knowledge discovery

Inductive logic programming (ILP) is concerned with the induction of logic programs from examples and background knowledge. In ILP, the shift of attention from program synthesis to knowledge discovery resulted in advanced techniques that are practically applicable for discovering knowledge in relational databases. This paper gives a brief introduction to ILP, presents selected ILP techniques for relational knowledge discovery and reviews selected ILP applications. Nada Lavrač, Ph.D.: She is a senior research associate at the Department of Intelligent Systems, J. Stefan Institute, Ljubljana, Slovenia (since 1978) and a visiting professor at the Klagenfurt University, Austria (since 1987). Her main research interest is in machine learning, in particular inductive logic programming and intelligent data analysis in medicine. She received a BSc in Technical Mathematics and MSc in Computer Science from Ljubljana University, and a PhD in Technical Sciences from Maribor University, Slovenia. She is coauthor of KARDIO: A Study in Deep and Qualitative Knowledge for Expert Systems, The MIT Press 1989, and Inductive Logic Programming: Techniques and Applications, Ellis Horwood 1994, and coeditor of Intelligent Data Analysis in Medicine and Pharmacology, Kluwer 1997. She was the coordinator of the European Scientific Network in Inductive Logic Programming ILPNET (1993–1996) and program cochair of the 8th European Machine Learning Conference ECML’95, and 7th International Workshop on Inductive Logic Programming ILP’97. Sašo Džeroski, Ph.D.: He is a research associate at the Department of Intelligent Systems, J. Stefan Institute, Ljubljana, Slovenia (since 1989). He has held visiting researcher positions at the Turing Institute, Glasgow (UK), Katholieke Universiteit Leuven (Belgium), German National Research Center for Computer Science (GMD), Sankt Augustin (Germany) and the Foundation for Research and Technology-Hellas (FORTH), Heraklion (Greece). His research interest is in machine learning and knowledge discovery in databases, in particular inductive logic programming and its applications and knowledge discovery in environmental databases. He is co-author of Inductive Logic Programming: Techniques and Applications, Ellis Horwood 1994. He is the scientific coordinator of ILPnet2, The Network of Excellence in Inductive Logic Programming. He was program co-chair of the 7th International Workshop on Inductive Logic Programming ILP’97 and will be program co-chair of the 16th International Conference on Machine Learning ICML’99. Masayuki Numao, Ph.D.: He is an associate professor at the Department of Computer Science, Tokyo Institute of Technology. He received a bachelor of engineering in electrical and electronics engineering in 1982 and his Ph.D. in computer science in 1987 from Tokyo Institute of Technology. He was a visiting scholar at CSLI, Stanford University from 1989 to 1990. His research interests include Artificial Intelligence, Global Intelligence and Machine Learning. Numao is a member of Information Processing Society of Japan, Japanese Society for Artificial Intelligence, Japanese Cognitive Science Society, Japan Society for Software Science and Technology and AAAI.     

Mining negative generalized knowledge from relational databases

Attribute-oriented induction (AOI) is a useful data mining method for extracting generalized knowledge from relational data and users’ background knowledge. Concept hierarchies can be integrated with the AOI method to induce multi-level generalized knowledge. However, the existing AOI approaches are only capable of mining positive knowledge from databases; thus, rare but important negative generalized knowledge that is unknown, unexpected, or contradictory to what the user believes, can be missed. In this study, we propose a global negative attribute-oriented induction (GNAOI) approach that can generate comprehensive and multiple-level negative generalized knowledge at the same time. Two pruning properties, the downward level closure property and the upward superset closure property, are employed to improve the efficiency of the algorithm, and a new interest measure, nim(cl), is exploited to measure the degree of the negative relation. Experiment results from a real-life dataset show that the proposed method is effective in finding global negative generalized knowledge.     

An approach to capturing and reusing tacit design knowledge using relational learning for knowledge graphs

Tacit design knowledge plays an important role in the process of product design and is a valuable knowledge asset for enterprises. In terms of the characteristics of tacit rational design knowledge, this paper puts forward a scientific hypothesis and approach on capturing and reusing tacit rational design knowledge. The presented approach represents the observable design result facts of products using design knowledge graphs. A design issue-solving oriented knowledge graph model is presented, where directed relation edges represent design issues, and nodes stand for design solutions. When a new design solutions requirement needs to be searched, tacit design knowledge can be reused by relational learning for the constructed design knowledge graphs. In relational learning, the design knowledge graph is converted into a three-order tensor, where two modes are solution nodes, and the third mode holds the issue relations. Then, a tensor factorization approach is employed to calculate the latent features between design solutions for an issue relation. As a result, a score vector to represent the existence of issue-solution relations can be obtained. By sorting the scores in descending order, we may select the solution node with the highest score as the design solution to be searched. Finally, a stamping die design case study is provided. The case study shows that the proposed approach is feasible, and effective, and has better flexibility, scalability and efficiency than CBR methods.     

基于模糊关联聚类的知识网选择方法

针对知识化制造系统中相似知识网日益增多和用户需求表达不清晰等导致的知识网选择问题,提出一种基于模糊关联聚类的知识网选择方法.综合知识网功能、完善程度和结构等方面构造的相似度具有反映知识网运算规律的特征.将两两知识网的相似度作为聚类数据,降低了高维特征空间的维数.模糊关联矩阵的分解,获得了知识网-类关系.目标知识网与类中类隶属度高的知识网的比较缩小了用户选择范围.最后的实例表明该方法是有效可行的.     

Associative learning using similarity knowledge bases for relational database search

A knowledge-based interface system for relational databases is presented. The system consists of two basic components; a similarity knowledge base and a generalization reasoning engine. The knowledge base is a particular type of semantic network that represents similarity relationships between the data. The similarity relationships are derived from the relational database by a psychological model of similarity. The reasoning engine propagates information obtained from the user's responses over the similarity network and also determines the generic kind of question it should next pose to the user. Through this dialog process, the interface system aids the user in specifying requests for relevant data, as well as in retrieving data from the database. Following a presentation of the constructs that support database design, we note that agenda dependency of the dialog results will typically occur. A smoothing mechanism that can reduce this dependency is presented. Incorporation of smoothing in the database design improves the efficiency and the effectiveness of the resulting system.     

基于关系记忆的胶囊网络知识图谱嵌入模型

作为一种语义知识库,知识图谱（KG）使用结构化三元组的形式存储真实世界的实体及其内在关系。为了推理知识图谱中缺失的真实三元组,考虑关系记忆网络较强的三元组表征能力和胶囊网络强大的特征处理能力,提出一种基于关系记忆的胶囊网络知识图谱嵌入模型。首先,通过编码实体和关系之间的潜在依赖关系和部分重要信息形成编码嵌入向量;然后,把嵌入向量与过滤器卷积以生成不同的特征图,再重组为对应的胶囊;最后,通过压缩函数和动态路由指定从父胶囊到子胶囊的连接,并根据子胶囊与权重内积的得分判断当前三元组的可信度。链接预测实验的结果表明,与CapsE模型相比,在倒数平均排名（MRR）和Hit@10评价指标上,所提模型在WN18RR数据集上分别提高了7.95%和2.2个百分点,在FB15K-237数据集上分别提高了3.82%和2个百分点。实验结果表明,所提模型可以更准确地推断出头实体和尾实体之间的关系。     

On the importance of relational concept knowledge in referral networks

This paper deals with the topic of peer-to-peer referral systems and the policies that allow for the emergence of efficient retrieval of requested information. In an agent-based peer-to-peer network, member agents are capable of giving and following referrals to each other. This results in the emergence of communities where agents directly interact with other neighboring agents that supply the required service or will refer the right source. The notion of referral networks, as presented in the work of (Yolum and Singh, Proceedings of the Second International Joint Conference on Autonomous Agents and Multiagent Systems, pp 592–599, 2003), and their application to knowledge management, lacks two fundamental aspects; the relation of concepts within a domain, and the ability of an agent to dynamically change their interests based on suggestions in the form of concept relations. This paper introduces the concept of an oracle agent, which is an agent with relational concept knowledge that can supply suggestions to a querying agent on how to adapt their interests. Additionally the notion of health and localized trust automata are used to aid agents in discriminating useful pieces of concept knowledge. These new features allow agents to search in new ways so as to achieve superior results and as a consequence outperform agents in a traditional referral network. The paper presents simulation results that confirm this hypothesis.     

存储于关系数据库的知识库的语义推理方法

语义推理的功能使得知识库更具人工智能,具有实用意义。文章根据语义模型的特点,构建了基于关系数据库的知识库语义存储体系,该存储体系的模式空间和实例空间相分离,降低了结构和数据的耦合性,使语义的存储范围更具完整性、语义的推理效果更具智能性。面向该存储体系的语义推理方法实现了相关语义(专家知识)的推理和相似语义(词汇)的推理,同时,该方法也考虑到了推理范围的可控能力和推理结果的语义还原能力。分析表明,该方法能应用于实际,但仍存在一些可改进之处。     

基于先验信息和谱分析的聚类融合算法

在聚类过程中利用先验信息能显著提高聚类算法的性能,但已存在的聚类融合算法很少考虑到数据集的先验信息。基于先验信息和谱分析,提出一种聚类融合算法,将成对限制信息引入到谱聚类算法中,用受限的谱聚类算法产生聚类成员,再采用基于互联合矩阵的集成方法生成最后的聚类结果。实验结果表明,利用先验信息能有效提高聚类的效果。     

Ontology extraction from relational database: Concept hierarchy as background knowledge

Relational Database (RDB) has been widely used as the back-end database of information system. Contains a wealth of high-quality information, RDB provides conceptual model and metadata needed in the ontology construction. However, most of the existing ontology building approaches convert RDB schema without considering the knowledge resided in the database. This paper proposed the approach for ontology extraction on top of RDB by incorporating concept hierarchy as background knowledge. Incorporating the background knowledge in the building process of Web Ontology Language (OWL) ontology gives two main advantages: (1) accelerate the building process, thereby minimizing the conversion cost; (2) background knowledge guides the extraction of knowledge resided in database. The experimental simulation using a gold standard shows that the Taxonomic F-measure (TF) evaluation reaches 90% while Relation Overlap (RO) is 83.33%. In term of processing time, this approach is more efficient than the current approaches. In addition, our approach can be applied in any of the fields such as eGoverment, eCommerce and so on.