Search

With the proliferation of knowledge graphs, modeling data with complex multirelational structure has gained increasing attention in the area of statistical relational learning. One of the most important goals of statistical relational learning is link prediction, i.e., predicting whether certain relations exist in the knowledge graph. A large number of models and algorithms have been proposed to perform link prediction, among which tensor factorization method has proven to achieve state-of-the-art performance in terms of computation efficiency and prediction accuracy. However, a common drawback of the existing tensor factorization models is that the missing relations and non-existing relations are treated in the same way, which results in a loss of information. To address this issue, we propose a binary tensor factorization model with probit link, which not only inherits the computation efficiency from the classic tensor factorization model but also accounts for the binary nature of relational data. Our proposed probit tensor factorization (PTF) model shows advantages in both the prediction accuracy and interpretability, 30 pages

Gaussian Process (GP) offers a principled non-parametric framework for learning stochastic functions. The generalization capability of GPs depends heavily on the kernel function, which implicitly imposes the smoothness assumptions of the data. However, common feature-based kernel functions are inefficient to model the relational data, where the smoothness assumptions implied by the kernels are violated. To model the complex and non-differentiable functions over relational data, we propose a novel Graph Convolutional Kernel, which enables to incorporate relational structures to feature-based kernels to capture the statistical structure of data. To validate the effectiveness of proposed kernel function in modeling relational data, we introduce GP models with Graph Convolutional Kernel in two relational learning settings, i.e., unsupervised settings of link prediction and semi-supervised settings of object classification. The parameters of our GP models are optimized through the scalable variational inducing point method. However, the highly structured likelihood objective requires densely sampling from variational distributions, which is costly and makes its optimization challenging in the unsupervised settings. To tackle this challenge, we propose a Local Neighbor Sampling technique with a provable more efficient computational complexity. Experimental results on real-world datasets demonstrate that our model achieves state-of-the-art performance in two relational learning tasks.

Recommender systems have played a critical role in many web applications to meet user's personalized interests and alleviate the information overload. In this survey, we review the development of recommendation frameworks with the focus on heterogeneous relational learning, which consists of different types of dependencies among users and items. The objective of this task is to map heterogeneous relational data into latent representation space, such that the structural and relational properties from both user and item domain can be well preserved. To address this problem, recent research developments can fall into three major lines: social recommendation, knowledge graph-enhanced recommender system, and multi-behavior recommendation. We discuss the learning approaches in each category, such as matrix factorization, attention mechanism and graph neural networks, for effectively distilling heterogeneous contextual information. Finally, we present an exploratory outlook to highlight several promising directions and opportunities in heterogeneous relational learning frameworks for recommendation., Comment: Published as a paper in IJCAI 2021

A large body of machine learning and AI is focused on learning models composed of (probabilistic) logical rules, i.e., relational models, over relational databases and knowledge bases. To learn effective relational models over the huge space of possible ones efficiently, users of the current learning systems must restrict the structure of the candidate models using language bias. ML experts have to spend a long time inspecting the data and performing many rounds of trial and error to develop an effective language bias. We propose AutoBias, a system that leverages information in the underlying data to generate the language bias. As its induced language bias may not restrict the set of candidate models as tightly as the manually-written ones, learning may not scale to large datasets. Thus, we design novel and efficient methods to sample and learn effective relational models over large data. Our extensive empirical study shows that AutoBias delivers the same accuracy as using manually-written language bias by imposing only a slight overhead on the learning time.

Handling relational data streams has become a crucial task, given the availability of pervasive sensors and Internet-produced content, such as social networks and knowledge graphs. In a relational environment, this is a particularly challenging task, since one cannot assure that the streams of examples are independent along the iterations. Thus, most relational learning systems are still designed to learn only from closed batches of data. Furthermore, in case there is a previously acquired model, these systems either would discard it or assuming it as correct. In this work, we propose an online relational learning algorithm that can handle continuous, open-ended streams of relational examples as they arrive. We employ techniques of theory revision to take advantage of the previously acquired model as a starting point, by finding where it should be modified to cope with the new examples, and automatically update it. We rely on the Hoeffding’s bound statistical theory to decide if the model must, in fact, be updated in accordance with the new examples. The proposed algorithm is built upon ProPPR statistical relational language, aiming at contemplating the uncertainty inherent to real data. Experimental results in social networks and entity co-reference datasets show the potential of the proposed approach compared to other relational learners.

Defining appropriate distance functions is a crucial aspect of effective and efficient similarity-based prediction and retrieval. Relational data are especially challenging in this regard. By viewing relational data as multi-relational graphs, one can easily see that a distance between a pair of nodes can be defined in terms of a virtually unlimited class of features, including node attributes, attributes of node neighbors, structural aspects of the node neighborhood and arbitrary combinations of these properties. In this paper we propose a rich and flexible class of metrics on graph entities based on earth mover’s distance applied to a hierarchy of complex counts-of-counts statistics. We further propose an approximate version of the distance using sums of marginal earth mover’s distances. We show that the approximation is correct for many cases of practical interest and allows efficient nearest-neighbor retrieval when combined with a simple metric tree data structure. An experimental evaluation on two real-world scenarios highlights the flexibility of our framework for designing metrics representing different notions of similarity. Substantial improvements in similarity-based prediction are reported when compared to solutions based on state-of-the-art graph kernels.

Dealing with relational learning generally relies on tools modeling relational data. An undirected graph can represent these data with vertices depicting entities and edges describing the relationships between the entities. These relationships can be well represented by multiple undirected graphs over the same set of vertices with edges arising from different graphs catching heterogeneous relations. The vertices of those networks are often structured in unknown clusters with varying properties of connectivity. These multiple graphs can be structured as a three-way tensor, where each slice of tensor depicts a graph which is represented by a count data matrix. To extract relevant clusters, we propose an appropriate model-based co-clustering capable of dealing with multiple graphs. The proposed model can be seen as a suitable tensor extension of mixture models of graphs, while the obtained co-clustering can be treated as a consensus clustering of nodes from multiple graphs. Applications on real datasets and comparisons with multi-view clustering and tensor decomposition methods show the interest of our contribution., Bundesministerium für Wirtschaft und Energie, Projekt DEAL

Relational learning addresses the task of learning models or patterns from relational data. Complementary to relational learning approaches that learn directly from relational data, developed in the Inductive Logic Programming research community, this chapter addresses the propositionalization approach of first transforming a relational database into a single-table representation, followed by a model or pattern construction step using a standard machine learning algorithm.

This chapter provides an introduction to standard machine learning approaches that learn from tabular data representations, followed by an outline of approaches using various other data types addressed in this monograph: texts, relational databases, and networks (graphs, knowledge graphs, and ontologies). We first briefly sketch the historical outline of the research area, establish the basic terminology, and categorize learning tasks in Sect. 2.1. Section 2.2 provides a short introduction to text mining. Section 2.3 introduces relational learning techniques, followed by a brief introduction to network analysis, including semantic data mining, in Sect. 2.4. The means for evaluating the performance of machine learning algorithms, when used for prediction and rule quality estimation, are outlined in Sect. 2.5. We outline selected data mining techniques and platforms in Sect. 2.6. Finally, Sect. 2.7 presents the implemented software that allows for running selected methods on illustrative examples.

Knowledge graphs became a popular means for modeling complex biological systems where they model the interactions between biological entities and their effects on the biological system. They also provide support for relational learning models which are known to provide highly scalable and accurate predictions of associations between biological entities. Despite the success of the combination of biological knowledge graph and relation learning models in biological predictive tasks, there is a lack of unified biological knowledge graph resources. This forced all current efforts and studies for applying a relational learning model on biological data to compile and build biological knowledge graphs from open biological databases. This process is often performed inconsistently across such efforts, especially in terms of choosing the original resources, aligning identifiers of the different databases, and assessing the quality of included data. To make relational learning on biomedical data more standardised and reproducible, we propose a new biological knowledge graph which provides a compilation of curated relational data from open biological databases in a unified format with common, interlinked identifiers. We also provide a new module for mapping identifiers and labels from different databases which can be used to align our knowledge graph with biological data from other heterogeneous sources. Finally, to illustrate the practical relevance of our work, we provide a set of benchmarks based on the presented data that can be used to train and assess the relational learning models in various tasks related to pathway and drug discovery.

Statistical machine learning algorithms usually assume that there is considerably-size data to train the models. However, they would fail in addressing domains where data is difficult or expensive to obtain. Transfer learning has emerged to address this problem of learning from scarce data by relying on a model learned in a source domain where data is easy to obtain to be a starting point for the target domain. On the other hand, real-world data contains objects and their relations, usually gathered from noisy environment. Finding patterns through such uncertain relational data has been the focus of the Statistical Relational Learning (SRL) area. Thus, to address domains with scarce, relational, and uncertain data, in this paper, we propose TreeBoostler, an algorithm that transfers the SRL state-of-the-art Boosted Relational Dependency Networks learned in a source domain to the target domain. TreeBoostler first finds a mapping between pairs of predicates to accommodate the additive trees into the target vocabulary. After, it employs two theory revision operators devised to handle incorrect relational regression trees aiming at improving the performance of the mapped trees. In the experiments presented in this paper, TreeBoostler has successfully transferred knowledge among several distinct domains. Moreover, it performs comparably or better than learning from scratch methods in terms of accuracy and outperforms a transfer learning approach in terms of accuracy and runtime.

Causal inference is at the heart of empirical research in natural and social sciences and is critical for scientific discovery and informed decision making. The gold standard in causal inference is performing randomized controlled trials ; unfortunately these are not always feasible due to ethical, legal, or cost constraints. As an alternative, methodologies for causal inference from observational data have been developed in statistical studies and social sciences. However, existing methods critically rely on restrictive assumptions such as the study population consisting of homogeneous elements that can be represented in a single flat table, where each row is referred to as a unit. In contrast, in many real-world settings, the study domain naturally consists of heterogeneous elements with complex relational structure, where the data is naturally represented in multiple related tables. In this paper, we present a formal framework for causal inference from such relational data. We propose a declarative language called CARL for capturing causal background knowledge and assumptions, and specifying causal queries using simple Datalog-like rules. CARL provides a foundation for inferring causality and reasoning about the effect of complex interventions in relational domains. We present an extensive experimental evaluation on real relational data to illustrate the applicability of CARL in social sciences and healthcare.

Statistical relational learning (SRL) is a field of machine learning that enables effective and robust reasoning about relational data structures. Several conventional data mining methods have been adapted for direct application to relational data representation. Probabilistic relational models extend Bayesian networks to a relational data mining context. To use this model, it is first necessary to build it: The structure and parameters of a probabilistic relational model must be set manually or learned from a relational observational dataset. Learning the structure remains the most complicated issue as it is a NP-hard problem. Existing approaches for probabilistic relational models structure learning are inspired from classical methods of learning the structure of Bayesian networks. The evaluation of learning approaches requires testing datasets and evaluation measurements. Processes to randomly generate the model and the data are already established in the relational context. However, metrics to evaluate a probabilistic relational model structure learning algorithm are not yet proposed. In fact, only the classical Recall and Precision have been used as evaluation metrics. In this work, we discuss why these metrics are not appropriate in this context, and we propose an adaptation of them to fit probabilistic relational models representation.

A generative probabilistic model for relational data consists of a family of probability distributions for relational structures over domains of different sizes. In most existing statistical relational learning (SRL) frameworks, these models are not projective in the sense that the marginal of the distribution for size-$n$ structures on induced sub-structures of size $k, Comment: Extended version (with proof appendix) of paper that is too appear in Proceedings of IJCAI 2020

Nowadays large scale Knowledge Bases (KBs) represent very important resources when it comes to develop expert systems. However, despite their huge sizes, KBs often suffer from incompleteness. Recently, much effort has been devoted in developing learning models to reduce the aforementioned issue.

We demonstrate a declarative differentiable programming framework based on the language of Lifted Relational Neural Networks, where small parameterized logic programs are used to encode relational learning scenarios. When presented with relational data, such as various forms of graphs, the program interpreter dynamically unfolds differentiable computational graphs to be used for the program parameter optimization by standard means. Following from the used declarative Datalog abstraction, this results into compact and elegant learning programs, in contrast with the existing procedural approaches operating directly on the computational graph level. We illustrate how this idea can be used for an efficient encoding of a diverse range of existing advanced neural architectures, with a particular focus on Graph Neural Networks (GNNs). Additionally, we show how the contemporary GNN models can be easily extended towards higher relational expressiveness. In the experiments, we demonstrate correctness and computation efficiency through comparison against specialized GNN deep learning frameworks, while shedding some light on the learning performance of existing GNN models., Comment: Submitted to MLJ's Special Track on Learning and Reasoning (May 15th 2020 cut-off) http://lr2020.iit.demokritos.gr/

Given a relational database and training examples for a target relation, relational learning algorithms learn a Datalog program that defines the target relation in terms of the existing relations in the database. We demonstrate CastorX, a relational learning system that performs relational learning over heterogeneous databases. The user specifies matching attributes between (heterogeneous) databases through matching dependencies. Because the content in these attributes may not match exactly, CastorX uses similarity operators to find matching values in these attributes. As the learning process may become expensive, CastorX implements sampling techniques that allow it to learn efficiently and output accurate definitions.

Relational program verification is a variant of program verification where one can reason about two programs and as a special case about two executions of a single program on different inputs. Relational program verification can be used for reasoning about a broad range of properties, including equivalence and refinement, and specialized notions such as continuity, information flow security or relative cost. In a higher-order setting, relational program verification can be achieved using relational refinement type systems, a form of refinement types where assertions have a relational interpretation. Relational refinement type systems excel at relating structurally equivalent terms but provide limited support for relating terms with very different structures. We present a logic, called Relational Higher Order Logic (RHOL), for proving relational properties of a simply typed $\lambda$-calculus with inductive types and recursive definitions. RHOL retains the type-directed flavour of relational refinement type systems but achieves greater expressivity through rules which simultaneously reason about the two terms as well as rules which only contemplate one of the two terms. We show that RHOL has strong foundations, by proving an equivalence with higher-order logic (HOL), and leverage this equivalence to derive key meta-theoretical properties: subject reduction, admissibility of a transitivity rule and set-theoretical soundness. Moreover, we define sound embeddings for several existing relational type systems such as relational refinement types and type systems for dependency analysis and relative cost, and we verify examples that were out of reach of prior work., Comment: Submitted to ICFP 2017

We present a new statistical relational learning (SRL) framework that supports reasoning with soft quantifiers, such as “most” and “a few.” We define the syntax and the semantics of this language, which we call $$\hbox {PSL}^Q$$ , and present a most probable explanation inference algorithm for it. To the best of our knowledge, $$\hbox {PSL}^Q$$ is the first SRL framework that combines soft quantifiers with first-order logic rules for modelling uncertain relational data. Our experimental results for two real-world applications, link prediction in social trust networks and user profiling in social networks, demonstrate that the use of soft quantifiers not only allows for a natural and intuitive formulation of domain knowledge, but also improves inference accuracy.

Clustering is an underspecified task: there are no universal criteria for what makes a good clustering. This is especially true for relational data, where similarity can be based on the features of individuals, the relationships between them, or a mix of both. Existing methods for relational clustering have strong and often implicit biases in this respect. In this paper, we introduce a novel similarity measure for relational data. It is the first measure to incorporate a wide variety of types of similarity, including similarity of attributes, similarity of relational context, and proximity in a hypergraph. We experimentally evaluate how using this similarity affects the quality of clustering on very different types of datasets. The experiments demonstrate that (a) using this similarity in standard clustering methods consistently gives good results, whereas other measures work well only on datasets that match their bias; and (b) on most datasets, the novel similarity outperforms even the best among the existing ones., 9 pages, 3 figures, 4 tables, submitted to ECMLPKDD 2017

To address the increasing relational challenges in international R&D collaboration, the present study develops a framework for understanding retrospective relational sensemaking in R&D offshore relationships. Using a comparative case study methodology, this study analyzes relational data from 56 interviews regarding four R&D offshore relationships between two large Swedish multinational companies and four R&D offshore partners. This study contributes to existing sensemaking theory by constructing a framework for retrospective relational sensemaking, including triggers and the phases of enactment, selection, and retention, to improve relational learning in R&D offshore relationships.

Traditional data mining algorithms will not work efficiently for most of the real world applications where the data is stored in relational format. Useful patterns can certainly be extracted from multiple relations using an existing traditional learning algorithm of data mining, but it would involve a lot of complexity. So there is a need of a multi relational classification, which analyzes relational data and predicts unknown patterns automatically. Moreover the performances of existing relational classifiers are limited, because the existing algorithms are not able to use different classifiers based on characteristics of different relations. The goal of the proposed approach is to select appropriate classifiers based on characteristics of different relations in the relational database to improve the overall performance without affecting the running time. So multi criteria classifier selection function based on ratio of accuracy and running time is used to select the most efficient classifier usi...

Deep learning has consistently defied state-of-the-art techniques in many fields over the last decade. However, we are just beginning to understand the capabilities of neural learning in symbolic domains. Deep learning architectures that employ parameter sharing over graphs can produce models which can be trained on complex properties of relational data. These include highly relevant NP-Complete problems, such as SAT and TSP. In this work, we showcase how Graph Neural Networks (GNN) can be engineered - with a very simple architecture - to solve the fundamental combinatorial problem of graph colouring. Our results show that the model, which achieves high accuracy upon training on random instances, is able to generalise to graph distributions different from those seen at training time. Further, it performs better than the Neurosat, Tabucol and greedy baselines for some distributions. In addition, we show how vertex embeddings can be clustered in multidimensional spaces to yield constructive solutions even though our model is only trained as a binary classifier. In summary, our results contribute to shorten the gap in our understanding of the algorithms learned by GNNs, as well as hoarding empirical evidence for their capability on hard combinatorial problems. Our results thus contribute to the standing challenge of integrating robust learning and symbolic reasoning in Deep Learning systems.

Multi-Entity Bayesian Network (MEBN) is a knowledge representation formalism combining Bayesian Networks (BNs) with First-Order Logic (FOL). MEBN has sufficient expressive power for general-purpose knowledge representation and reasoning, and is the logical basis of Probabilistic Web Ontology Language (PR-OWL), a representation language for probabilistic ontologies. Developing an MEBN model to support a given application is a challenge, requiring definition of entities, relationships, random variables, conditional dependence relationships, and probability distributions. When available, data can be invaluable both to improve performance and to streamline development. By far the most common format for available data is the relational database (RDB). Relational databases describe and organize data according to the Relational Model (RM). Developing an MEBN model from data stored in an RDB therefore requires mapping between the two formalisms. This paper presents MEBN-RM, a set of mapping rules between key elements of MEBN and RM. We identify links between the two languages (RM and MEBN) and define four levels of mapping from elements of RM to elements of MEBN. These definitions are implemented in the MEBN-RM algorithm, which converts a relational schema in RM to a partial MEBN model. Through this research, the software has been released as an MEBN-RM open-source software tool. The method is illustrated through two example use cases using MEBN-RM to develop MEBN models: a Critical Infrastructure Defense System and a Smart Manufacturing System. Both systems are proof-of-concept systems used for situation awareness, where data coming from various sensors are stored in RDBs and converted into MEBN models through the MEBN-RM algorithm. In these use cases, we evaluate the performance of the MEBN-RM algorithm in terms of mapping speed and quality to show its efficiency in MEBN modeling.

In many applications of relational learning, the available data can be seen as a sample from a larger relational structure (e.g. we may be given a small fragment from some social network). In this paper we are particularly concerned with scenarios in which we can assume that (i) the domain elements appearing in the given sample have been uniformly sampled without replacement from the (unknown) full domain and (ii) the sample is complete for these domain elements (i.e. it is the full substructure induced by these elements). Within this setting, we study bounds on the error of sufficient statistics of relational models that are estimated on the available data. As our main result, we prove a bound based on a variant of the Vapnik-Chervonenkis dimension which is suitable for relational data., Longer version of paper accepted at ECML PKDD 2018

Due to the growing amount of data generated and stored in relational databases, relational learning has attracted the interest of researchers in recent years. Many approaches have been developed in order to learn relational data. One of the approaches used to learn relational data is Dynamic Aggregation of Relational Attributes (DARA). The DARA algorithm is designed to summarize relational data with one-to-many relations. However, DARA suffers a major drawback when the cardinalities of attributes are very high because the size of the vector space representation depends on the number of unique values that exist for all attributes in the dataset. A feature selection process can be introduced to overcome this problem. These selected features can then be further optimized to achieve a good classification result when using k Nearest Neighbour (k-NN) classifier. Several clustering runs can be performed for different values of k to yield an ensemble of clustering results. This paper proposes a two-layered genetic algorithm-based feature selection in order to improve the classification performance of learning relational database using a k-NN ensemble classifier. The proposed method involves the task of omitting less relevant features but retaining the diversity of the classifiers so as to improve the performance of the k-NN ensemble. Based on the results obtained, it shows that the proposed k-NN ensemble is able to improve the performance of traditional k-NN classifiers.

This tutorial overviews the state of the art in learning models over relational databases and makes the case for a first-principles approach that exploits recent developments in database research. The input to learning classification and regression models is a training dataset defined by feature extraction queries over relational databases. The mainstream approach to learning over relational data is to materialize the training dataset, export it out of the database, and then learn over it using a statistical package. This approach can be expensive as it requires the materialization of the training dataset. An alternative approach is to cast the machine learning problem as a database problem by transforming the data-intensive component of the learning task into a batch of aggregates over the feature extraction query and by computing this batch directly over the input database. The tutorial highlights a variety of techniques developed by the database theory and systems communities to improve the performance of the learning task. They rely on structural properties of the relational data and of the feature extraction query, including algebraic (semi-ring), combinatorial (hypertree width), statistical (sampling), or geometric (distance) structure. They also rely on factorized computation, code specialization, query compilation, and parallelization., Comment: 10 pages, 1 figure

Despite the huge amount of work devoted to the treatment of time within the relational context, few relevant temporal phenomena still remain to be addressed. One of them is the treatment of “ nearly periodic events ”, i.e., events/facts that occur in intervals of time which repeat periodically (e.g., a meeting occurring twice each Monday , possibly not at regular times ). Nearly periodic events are quite frequent in everyday life, and thus in many applicative contexts. Their treatment within the relational model is quite challenging, since it involves the integrated treatment of three aspects: (i) the number of repetitions, (ii) their periodicity , and (iii) temporal indeterminacy . Coping with this problem requires an in-depth extension of current temporal relational database techniques. In this paper, we introduce a new data model, and new definitions of relational algebraic operators coping with the above issues. We ascertain the properties of the new model and algebra, with emphasis on the expressiveness of our representation model, on the reducibility property, and on the correctness of the algebraic operators.

Learning novel relations from relational databases is an important problem with many applications in database systems and machine learning. Relational learning algorithms leverage the properties of the database schema to find the definition of the target relation in terms of the existing relations in the database. However, the same data set may be represented under different schemas for various reasons, such as efficiency and data quality. Unfortunately, current relational learning algorithms tend to vary quite substantially over the choice of schema, which complicates their off-the-shelf application. We demonstrate Castor , a relational learning system that efficiently learns the same definitions over common schema variations. The results of Castor are more accurate than well-known learning systems over large data.

Compound approximation spaces for relational data are proposed.They are viewed as expansions of tolerance approximation spaces to a relational case.Properties of compound approximation spaces are investigated.Compound approximation spaces enable to derive new knowledge from relational data. Rough set theory provides a powerful tool for dealing with uncertainty in data. Application of variety of rough set models to mining data stored in a single table has been widely studied. However, analysis of data stored in a relational structure using rough sets is still an extensive research area. This paper proposes compound approximation spaces and their constrained versions that are intended for handling uncertainty in relational data. The proposed spaces are expansions of tolerance approximation ones to a relational case. Compared with compound approximation spaces, the constrained version enables to derive new knowledge from relational data. The proposed approach can improve mining relational data that is uncertain, incomplete, or inconsistent.

Graphical abstractDisplay Omitted HighlightsFormula-based relations defined in a granular computing framework are proposed.The relations are used to construct more informative granules.The granules are used to represent relational data and patterns.Thanks to this approach, the generation of patterns can be speeded up.The approach makes it possible to discover richer knowledge from relational data. The complex structure of relational data makes the process of knowledge discovery from data a more challenging task compared with the single table data structure. The usefulness of granular computing based approaches to mining data stored in a single table is a driving force for adapting this method to relational data. This paper proposes relation-based granules that are defined in a granular computing based approach to mining relational data. The relations are used to represent relational data and patterns to be discovered. Thanks to this representation, the generation of patterns can be speeded up. The representation also makes it possible to discover richer knowledge from relational data.

Given a relational database and training examples for a target relation, relational learning algorithms learn a definition for the target relation in terms of the existing relations in the database. We propose a relational learning system called CastorX, which learns efficiently across multiple heterogeneous databases. The user specifies connections and relationships between different databases using a set of declarative constraints called matching dependencies (MDs). Each MD connects tuples across multiple databases that are related and can meaningfully join but the values of their join attributes may not be equal due to the different representations of these values in different databases. CastorX leverages these constraints during learning to find the information relevant to the training data and target definition across multiple databases. Since each tuple in a database may be connected to too many tuples in other databases according to an MD, the learning process will become very slow. Hence, CastorX uses sampling techniques to learn efficiently and output accurate definitions.

We consider the problem of learning Boltzmann machine classifiers from relational data. Our goal is to extend the deep belief framework of RBMs to statistical relational models. This allows one to exploit the feature hierarchies and the non-linearity inherent in RBMs over the rich representations used in statistical relational learning (SRL). Specifically, we use lifted random walks to generate features for predicates that are then used to construct the observed features in the RBM in a manner similar to Markov Logic Networks. We show empirically that this method of constructing an RBM is comparable or better than the state-of-the-art probabilistic relational learning algorithms on six relational domains.

During the last 35 years, data management principles such as physical and logical independence, declarative querying and cost-based optimization have led to profound pervasiveness of relational databases in any kind of organization. More importantly, these technical advances have enabled the first round of business intelligence applications and laid the foundation for managing and analyzing Big Data today.

We improved wordification methodology and provide a formal framework and pseudo code.We statistically evaluated comparable algorithms on multiple relational databases.Experiments show favorable results in terms of accuracy and efficiency.Feature simplicity is compensated by n-gram construction and by feature weighting.We implemented the full experimental workflow in a data mining platform ClowdFlows. Inductive Logic Programming (ILP) and Relational Data Mining (RDM) address the task of inducing models or patterns from multi-relational data. One of the established approaches to RDM is propositionalization, characterized by transforming a relational database into a single-table representation. This paper presents a propositionalization technique called wordification which can be seen as a transformation of a relational database into a corpus of text documents. Wordification constructs simple, easy to understand features, acting as words in the transformed Bag-Of-Words representation. This paper presents the wordification methodology, together with an experimental comparison of several propositionalization approaches on seven relational datasets. The main advantages of the approach are: simple implementation, accuracy comparable to competitive methods, and greater scalability, as it performs several times faster on all experimental databases. Furthermore, the wordification methodology and the evaluation procedure are implemented as executable workflows in the web-based data mining platform ClowdFlows. The implemented workflows include also several other ILP and RDM algorithms, as well as the utility components that were added to the platform to enable access to these techniques to a wider research audience.

Relational learning algorithms mine complex databases for interesting patterns. Usually, the search space of patterns grows very quickly with the increase in data size, making it impractical to solve important problems. In this work we present the design of a relational learning system, that takes advantage of graphics processing units (GPUs) to perform the most time consuming function of the learner, rule coverage. To evaluate performance, we use four applications: a widely used relational learning benchmark for predicting carcinogenesis in rodents, an application in chemo-informatics, an application in opinion mining, and an application in mining health record data. We compare results using a single and multiple CPUs in a multicore host and using the GPU version. Results show that the GPU version of the learner is up to eight times faster than the best CPU version.

We present a new approach lo learn from relational data based on re-representation of the examples. This approach, called property-based re-representation is based on a new analysis of the structure of refinement graphs used in ILP and relational learning in general. This analysis allows the characterization of relational examples by a set of multi-relational patterns called properties. Using them, we perform a property-based re-representation of relational examples that facilitates the development of relational learning techniques. Additionally, we show the usefulness of re-representation with a collection of experiments in the context of nearest neighbor classification. © 2015 - IOS Press and the authors., Research partially supported by Projects Next-CBR (TIN2009-13692-C03-01) and Cognitio (TIN2012-38450- C03-03) and by the Generalitat de Catalunya under the grants 2009-SGR-1434.

Information granulation is a powerful tool for data analysis and processing. However, not much attention has been devoted to application of this tool to data stored in a relational structure. This paper extends the notion of information granules to a relational case. Two information systems intended to store relational data are proposed. This study also extends a granule description language to express information granules derived from relational data. The proposed approach enables to analyze a given problem at different levels of granularity of relational data. This can find application in searching for patterns in data mining.

Multi-relational data (e.g., product design knowledge graph) learning has attracted great attention in the various research areas recently. Motivated by the assumption that the entities have appropriate relations with respect to their belonging categories (e.g., two entities in the {process parameter} and the {machine unit} categories respectively likely bear upon the relationship of {belong to}), this paper proposes a tensor factorization approach for multi-relational data in a supervised way from a Bayesian perspective, referred as supervised Bayesian tensor factorization (SetF). The proposed SetF is formulated as a joint optimization framework of probabilistic inference and e-insensitive support vector regression. The generative-discriminative nature of the proposed SetF is able to discover the latent representation of multi-relational data under the principle of the max-margin learning. The interplay between entities, relations, and the categories (entities involved) yields more predictive latent representations that are particularly appropriate for discriminant analysis. We conduct a set of experimental comparisons in terms of multi-relational learning tasks over several benchmark datasets, and also over antenna arrays design parameter relation analysis, which demonstrate the effectiveness of the proposed method.

Many applications, such as knowledge base completion and patient data, only have access to positive examples but lack negative examples which are required by standard ILP techniques and suffer under the closed-world assumption. The corresponding propositional problem is known as Positive and Unlabeled (PU) learning. In this field, it is known that using the label frequency (the fraction of true positive examples that are labeled) makes learning easier. This notion has not been explored yet in the relational domain. The goal of this work is twofold: 1) to explore if using the label frequency would also be useful when working with relational data and 2) to propose a method for estimating the label frequency from relational PU data. Our experiments confirm the usefulness of knowing the label frequency and of our estimate. ispartof: pages:16-30 ispartof: Proceedings of the 27th International Conference on Inductive Logic Programming (ILP) vol:10759 LNAI pages:16-30 ispartof: Inductive Logic Programming location:Orléans date:4 Sep - 6 Sep 2017 status: published

Relational databases are valuable resources for learning novel and interesting relations and concepts. Relational learning algorithms learn the definition of new relations in terms of the existing relations in the database. In order to constraint the search through the large space of candidate definitions, users must specify a language bias. Unfortunately, specifying the language bias is done via trial and error and is guided by the expert's intuitions. Hence, it normally takes a great deal of time and effort to effectively use these algorithms. We report our on-going work on building AutoMode, a system that leverages information in the schema and content of the database to automatically induce the language bias used by popular relational learning algorithms.

International audience; Probabilistic Relational Models (PRMs) extend Bayesian networks (BNs) with the notion of class of relational databases. Because of their richness, learning them is a difficult task. In this paper, we propose a method that learns a PRM from data using the semantic knowledge of an ontology describing these data in order to make the learning easier. To present our approach, we describe an implementation based on an on-tology of transformation processes and compare its performance to that of a method that learns a PRM directly from data. We show that, even with small datasets, our approach of learning a PRM using an ontology is more efficient.

International audience; Fault Trees or Bow Tie Diagrams are widely used for system dependability assessment. Some probabilistic extensions have been proposed by using Bayesian network formalism. This article proposes a general modeling approach under the form of a probabilistic relational model (PRM), relational extension of Bayesian networks, that can represent any fault tree, defined as an event tree with possible safety barriers, simply described in a relational database. We first describe an underlying relational schema describing a generic fault tree, and the probabilistic dependencies needed to model the existence of an event given the possible existence of its related causes and eventual safety barriers.

This chapter develops a general granular computing based framework for mining relational data. The framework provides a granular representation of relational data that is constructed based on target objects and the sets of background objects related to the target ones. The chapter also shows the possibility of deriving relational patterns from the granular representation.

Statistical Relational Learning (SRL) approaches have been developed to learn in presence of noisy relational data by combining probability theory with first order logic. While powerful, most learning approaches for these models do not scale well to large datasets. While advances have been made on using relational databases with SRL models [14], they have not been extended to handle the complex model learning (structure learning task). We present a scalable structure learning approach that combines the benefits of relational databases with search strategies that employ rich inductive bias from Inductive Logic Programming. We empirically show the benefits of our approach on boosted structure learning for Markov Logic Networks.

Probabilistic relational models (PRMs) extend the Bayesian network representation to incorporate a much richer relational structure. Existing probabilistic relational model (PRM) learning approaches based on search and scoring usually perform a heuristic search for the highest scoring structure. In this paper, we proposes the maximum likelihood tree based immune binary particle swarm optimization (MLT-IBPSO) method to learn structures of PRMs from relational data. First, a maximum likelihood tree (MLT) is generated from the data sample, and a population is created according to the MLT. Then, immune theory is combined with particle swarm optimization (PSO) for searching the structures. As a result, the probabilistic structure is learned based on the proposed method. Experiments show that the MLT-IBPSO method can learn structures from relational data effectively.

Valid-time indeterminacy is "don't know when" indeterminacy, coping with cases in which one does not exactly know when a fact holds in the modeled reality. In this paper, we first propose a reference representation (data model and algebra) in which all possible temporal scenarios induced by valid-time indeterminacy can be extensionally modeled. We then specify a family of 16 more compact representational data models. We demonstrate their correctness with respect to the reference representation and analyze several properties, including their data expressiveness. Then, we compare these compact models along several relevant dimensions. Finally, we also extend the reference representation and a representative of compact representations to cope with probabilities.

We introduce relational redescription mining, that is, the task of finding two structurally different patterns that describe nearly the same set of object pairs in a relational dataset. By extending redescription mining beyond propositional and real-valued attributes, it provides a powerful tool to match different relational descriptions of the same concept. We propose an alternating scheme for solving this problem. Its core consists of a novel relational query miner that efficiently identifies discriminative connection patterns between pairs of objects. Compared to a baseline Inductive Logic Programming (ILP) approach, our query miner is able to mine more complex queries, much faster. We performed extensive experiments on three real world relational datasets, and present examples of redescriptions found, exhibiting the power of the method to expressively capture relations present in these networks. ispartof: Machine Learning vol:96 issue:3 pages:225-248 status: published