Volumen 29 (2019): Heft 1 (March 2019)
Exploring Complex and Big Data (special section, pp. 7-91), Johann Gamper, Robert Wrembel (Eds.)

We present a preliminary study to introduce narratives as a first-class functionality in digital libraries. The general idea is to enrich those libraries with semantic networks of events providing a meaningful contextualisation of the digital libraries’ objects. More specific motivations are presented through a set of use cases by different actors who would benefit from using narratives for different purposes. Then, we consider a specific digital library, Europeana, the largest European digital library in the cultural heritage domain. We discuss how the Europeana Data Model could be extended for representing narratives, and we introduce an ontology for narratives. We also present a semi-automatic tool, which, on the basis of the ontology, supports the creation and visualisation of narratives, and we show how the tool has been employed to create a narrative of the life of the painter Gustav Klimt as a case study. In particular, we focus our attention on the functionality of the tool that allows extracting and proposing to the user specific digital objects for each event of the narrative.

Predictive analysis gradually gains importance in industry. For instance, service engineers at Siemens diagnostic centres unveil hidden knowledge in huge amounts of historical sensor data and use it to improve the predictive systems analysing live data. Currently, the analysis is usually done using data-dependent rules that are specific to individual sensors and equipment. This dependence poses significant challenges in rule authoring, reuse, and maintenance by engineers. One solution to this problem is to employ ontology-based data access (OBDA), which provides a conceptual view of data via an ontology. However, classical OBDA systems do not support access to temporal data and reasoning over it. To address this issue, we propose a framework for temporal OBDA. In this framework, we use extended mapping languages to extract information about temporal events in the RDF format, classical ontology and rule languages to reflect static information, as well as a temporal rule language to describe events. We also propose a SPARQL-based query language for retrieving temporal information and, finally, an architecture of system implementation extending the state-of-the-art OBDA platform Ontop.

In this paper, we study various ways of representing and querying fact data that are time-stamped with a time period in a data warehouse. The main focus is on how to represent the time periods that are associated with the facts in order to support convenient and efficient aggregations over time. We propose three distinct logical models that represent time periods as sets of all time points in a period (instant model), as pairs of start and end time points of a period (period model), and as atomic units that are explicitly stored in a new period dimension (period∗ model). The period dimension is enriched with information about the days of each period, thereby combining the former two models. We use four different classes of aggregation queries to analyze query formulation, query execution, and query performance over the three models. An extensive empirical evaluation on synthetic and real-world datasets and the analysis of the query execution plans reveal that the period model is the best choice in terms of runtime and space for all four query classes.

A prominent characteristic of clinical data is their heterogeneity—such data include structured examination records and laboratory results, unstructured clinical notes, raw and tagged images, and genomic data. This heterogeneity poses a formidable challenge while constructing diagnostic and therapeutic decision models that are currently based on single modalities and are not able to use data in different formats and structures. This limitation may be addressed using data fusion methods. In this paper, we describe a case study where we aimed at developing data fusion models that resulted in various therapeutic decision models for predicting the type of treatment (surgical vs. non-surgical) for patients with bone fractures. We considered six different approaches to integrate clinical data: one fusion model based on combination of data (COD) and five models based on combination of interpretation (COI). Experimental results showed that the decision model constructed following COI fusion models is more accurate than decision models employing COD. Moreover, statistical analysis using the one-way ANOVA test revealed that there were two groups of constructed decision models, each containing the set of three different models. The results highlighted that the behavior of models within a group can be similar, although it may vary between different groups.

Today’s ETL tools provide capabilities to develop custom code as user-defined functions (UDFs) to extend the expressiveness of the standard ETL operators. However, while this allows us to easily add new functionalities, it also comes with the risk that the custom code is not intended to be optimized, e.g., by parallelism, and for this reason, it performs poorly for data-intensive ETL workflows. In this paper we present a novel framework, which allows the ETL developer to choose a design pattern in order to write parallelizable code and generates a configuration for the UDFs to be executed in a distributed environment. This enables ETL developers with minimum expertise in distributed and parallel computing to develop UDFs without taking care of parallelization configurations and complexities. We perform experiments on large-scale datasets based on TPC-DS and BigBench. The results show that our approach significantly reduces the effort of ETL developers and at the same time generates efficient parallel configurations to support complex and data-intensive ETL tasks.

Finding clusters in high dimensional data is a challenging research problem. Subspace clustering algorithms aim to find clusters in all possible subspaces of the dataset, where a subspace is a subset of dimensions of the data. But the exponential increase in the number of subspaces with the dimensionality of data renders most of the algorithms inefficient as well as ineffective. Moreover, these algorithms have ingrained data dependency in the clustering process, which means that parallelization becomes difficult and inefficient. SUBSCALE is a recent subspace clustering algorithm which is scalable with the dimensions and contains independent processing steps which can be exploited through parallelism. In this paper, we aim to leverage the computational power of widely available multi-core processors to improve the runtime performance of the SUBSCALE algorithm. The experimental evaluation shows linear speedup. Moreover, we develop an approach using graphics processing units (GPUs) for fine-grained data parallelism to accelerate the computation further. First tests of the GPU implementation show very promising results.

The positivity and absolute stability of a class of fractional nonlinear continuous-time and discrete-time systems are addressed. Necessary and sufficient conditions for the positivity of this class of nonlinear systems are established. Sufficient conditions for the absolute stability of this class of fractional positive nonlinear systems are also given.

This paper presents the design of a state observer which guarantees quadratic boundedness of the estimation error. By using quadratic Lyapunov stability analysis, the convergence rate and the ultimate (steady-state) error bounding ellipsoid are identified as the parameters that define the behaviour of the estimation. Then, it is shown that these objectives can be merged in a scalarised objective function with one design parameter, making the design problem convex. In the second part of the article, a UAV model is presented which can be made linear by considering a particular state and frame of reference. The UAV model is extended to incorporate a disturbance model of variable size. The joint model matches the structure required to derive an observer, following the lines of the proposed design approach. An observer for disturbances acting on the UAV is derived and the analysis of the performances with respect to the design parameters is presented. The effectiveness and main characteristics of the proposed approach are shown using simulation results.

This paper considers reasonable bandwidth allocation for multiclass services in peer-to-peer (P2P) networks, measures the satisfaction of each peer as a customer by a utility function when acquiring one service, and develops an optimization model for bandwidth allocation with the objective of utility maximization. Elastic services with concave utilities are first considered and the exact expression of optimal bandwidth allocation for each peer is deduced. In order to obtain an optimum in distributed P2P networks, we develop a gradient-based bandwidth allocation scheme and illustrate the performance with numerical examples. Then we investigate bandwidth allocation for inelastic services with sigmoidal utilities, which is a nonconvex optimization problem. In order to solve it, we analyze provider capacity provisioning for bandwidth allocation of inelastic services and modify the update rule for prices that service customers should pay. Numerical examples are finally given to illustrate that the improved scheme can also efficiently converge to the global optimum.

We present two algorithms in which constrained spectral clustering is implemented as unconstrained spectral clustering on a multi-layer graph where constraints are represented as graph layers. By using the Nystrom approximation in one of the algorithms, we obtain time and memory complexities which are linear in the number of data points regardless of the number of constraints. Our algorithms achieve superior or comparative accuracy on real world data sets, compared with the existing state-of-the-art solutions. However, the complexity of these algorithms is squared with the number of vertices, while our technique, based on the Nyström approximation method, has linear time complexity. The proposed algorithms efficiently use both soft and hard constraints since the time complexity of the algorithms does not depend on the size of the set of constraints.

In this article, we discuss the implementation of a quantum recommendation system that uses a quantum variant of the k-nearest neighbours algorithm and the Grover algorithm to search for a specific element in an unstructured database. In addition to the presentation of the recommendation system as an algorithm, the article also shows the main steps in construction of a suitable quantum circuit for realisation of a given recommendation system. The computational complexity of individual calculation steps in the recommendation system is also indicated. The verification of the correctness of the proposed system is analysed as well, indicating an algebraic equation describing the probability of success of the recommendation. The article also shows numerical examples presenting the behaviour of the recommendation system for two selected cases.

Instance selection is often performed as one of the preprocessing methods which, along with feature selection, allows a significant reduction in computational complexity and an increase in prediction accuracy. So far, only few authors have considered ensembles of instance selection methods, while the ensembles of final predictive models attract many researchers. To bridge that gap, in this paper we compare four ensembles adapted to instance selection: Bagging, Feature Bagging, AdaBoost and Additive Noise. The last one is introduced for the first time in this paper. The study is based on empirical comparison performed on 43 datasets and 9 base instance selection methods. The experiments are divided into three scenarios. In the first one, evaluated on a single dataset, we demonstrate the influence of the ensembles on the compression–accuracy relation, in the second scenario the goal is to achieve the highest prediction accuracy, and in the third one both accuracy and the level of dataset compression constitute a multi-objective criterion. The obtained results indicate that ensembles of instance selection improve the base instance selection algorithms except for unstable methods such as CNN and IB3, which is achieved at the expense of compression. In the comparison, Bagging and AdaBoost lead in most of the scenarios. In the experiments we evaluate three classifiers: 1NN, kNN and SVM. We also note a deterioration in prediction accuracy for robust classifiers (kNN and SVM) trained on data filtered by any instance selection methods (including the ensembles) when compared with the results obtained when the entire training set was used to train these classifiers.

The focus of this research is to combine statistical and machine learning tools in application to a high-throughput biological data set on ionizing radiation response. The analyzed data consist of two gene expression sets obtained in studies of radiosensitive and radioresistant breast cancer patients undergoing radiotherapy. The data sets were similar in principle; however, the treatment dose differed. It is shown that introducing mathematical adjustments in data preprocessing, differentiation and trend testing, and classification, coupled with current biological knowledge, allows efficient data analysis and obtaining accurate results. The tools used to customize the analysis workflow were batch effect filtration with empirical Bayes models, identifying gene trends through the Jonckheere–Terpstra test and linear interpolation adjustment according to specific gene profiles for multiple random validation. The application of non-standard techniques enabled successful sample classification at the rate of 93.5% and the identification of potential biomarkers of radiation response in breast cancer, which were confirmed with an independent Monte Carlo feature selection approach and by literature references. This study shows that using customized analysis workflows is a necessary step towards novel discoveries in complex fields such as personalized individual therapy.

This paper deals with the synchronization of fractional-order chaotic discrete-time systems. First, some new concepts regarding the output-memory observability of non-linear fractional-order discrete-time systems are developed. A rank criterion for output-memory observability is derived. Second, a dead-beat observer which recovers exactly the true state system from the knowledge of a finite number of delayed inputs and delayed outputs is proposed. The case of the presence of an unknown input is also studied. Third, secure data communication based on a generalized fractional-order Hénon map is proposed. Numerical simulations and application to secure speech communication are presented to show the efficiency of the proposed approach.

High-order cumulant tensors carry information about statistics of non-normally distributed multivariate data. In this work we present a new efficient algorithm for calculation of cumulants of arbitrary orders in a sliding window for data streams. We show that this algorithm offers substantial speedups of cumulant updates compared with the current solutions. The proposed algorithm can be used for processing on-line high-frequency multivariate data and can find applications, e.g., in on-line signal filtering and classification of data streams. To present an application of this algorithm, we propose an estimator of non-Gaussianity of a data stream based on the norms of high order cumulant tensors. We show how to detect the transition from Gaussian distributed data to non-Gaussian ones in a data stream. In order to achieve high implementation efficiency of operations on super-symmetric tensors, such as cumulant tensors, we employ a block structure to store and calculate only one hyper-pyramid part of such tensors.