Neuroevolution of Augmenting Topologies (NEAT) and its advanced version, Evolvable-Substrate HyperNEAT (ES-HyperNEAT), have shown great potential in developing neural networks. However, their effectiveness heavily depends on the selection of hyperparameters. This study investigates the optimization of ES-HyperNEAT hyperparameters using the Tree-structured Parzen Estimator (TPE) on the MNIST classification task, exploring a search space of over 3 billion potential combinations. TPE effectively navigates this vast space, significantly outperforming random search in terms of mean, median, and best accuracy. During the validation process, the best hyperparameter configuration found by TPE achieves an accuracy of 29.00% on MNIST, surpassing previous studies while using a smaller population size and fewer generations. The transferability of the optimized hyperparameters is explored in logic operations and Fashion-MNIST tasks, revealing successful transfer to the more complex Fashion-MNIST problem but limited to simpler logic operations. This study emphasizes a method to unlock the full potential of neuroevolutionary algorithms and provides insights into the hyperparameters' transferability across tasks of varying complexity.

Arabic classical poetry meter is a sequence of patterns. A poetry verse is characterized by a meter and consists of two parts. Detecting classical poetry meter is important for teaching purposes, for poetry and prose categorization, for authorship recognition and for computational aesthetics. Automatically detecting the meter of any single verse written as a normal sentence is challenging. We need a global approach that processes phonological verse preparation, deals with distinguishing the first part of the verse from the second, handles meters disambiguation and covers the verse parts concordance. To tackle this challenge, we introduce a novel solution called the Arabic Meters Identification System (AMIS) that combines an exhaustive pattern data set, pattern matching and similarity. We evaluate our system on a vocalized poetry corpus and reach a precision of 99.3%.

Despite the many integration tools proposed for mapping between OWL ontologies and the object-oriented paradigm, developers are still reluctant to incorporate ontologies into their code repositories.In this paper we survey existing approaches for OWL to OOP mapping trying to identify reasons for this shy adoption of ontologies among conventional software developers.We present a classification of the surveyed approaches and tools based on the characteristics of their resulting artifacts.We finally provide our own reflection for other potential reasons beyond those addressed in the literature.

Most recent efforts on bringing ontologies into mainstream programming languages were hindered by some fundamental issues; mainly the lack of expressiveness of programming languages compared to the declarative nature of ontological languages as well as the different assumptions on which reasoning in these languages is based on.In this paper we give the idea of adopting ontological programming approaches a second thought by proposing a prototype for a C# ontological knowledgebase system where ontologies are expressed directly in an executable form.We present our experience on bridging the semantic gab in general purpose programming languages and on exploiting metaprogramming and the dynamic compilation feature of modern compilers for performing certain entailment queries without the need for the bulky ontology classification step usually required in the case of conventional ontological tools.

We introduce an end-to-end methodology (from text processing to querying a knowledge graph) for the sake of knowledge extraction from text corpora with a focus on a list of vocabularies of interest. We propose a pipeline that incorporates Natural Language Processing (NLP), Formal Concept Analysis (FCA), and Ontology Engineering techniques to build an ontology from textual data. We then extract the knowledge about controlled vocabularies by querying that knowledge graph, i.e., the engineered ontology. We demonstrate the significance of the proposed methodology by using it for knowledge extraction from a text corpus that consists of 800 news articles and reports about companies and products in the IT and pharmaceutical domain, where the focus is on a given list of 250 controlled vocabularies.

Ontologies have gone lengths in various areas of knowledge engineering, yet they are falling short of reaching an equal position as formal domain models in the landscape of enterprise software development. In this paper, we present an approach for integrating ontologies into the code space of conventional software. We argue that the limited adoption of ontologies in software development is partially due to the lack of imperative programming capabilities. We propose extending ontologies with procedural extensions by expressing them in an executable form. Finally, we discuss the advantages of this representation and the possibilities for further improvements.

Despite the many integration tools proposed for mapping between OWL ontologies and the object-oriented paradigm, developers are still reluctant to incorporate ontologies into their code repositories. In this paper we survey existing approaches for OWL-to-OOP mapping trying to identify reasons for this shy adoption of ontologies among conventional software developers. We present a classification of the surveyed approaches and tools based on their technical characteristics and their resulting artifacts. We discuss further potential reasons beyond what have been addressed in the literature before finally providing our own reflection and outlook.

Ontologies have been frequently used for representing domain knowledge. They have lots of applications in semantic knowledge extraction. However, learning ontologies especially from unstructured data is a difficult yet an interesting challenge. In this paper, we introduce a pipeline for learning ontology from a text corpus in a semi-automated fashion using Natural Language Processing (NLP) and Formal Concept Analysis (FCA). We apply our proposed method on a small given corpus that consists of some news documents in IT and pharmaceutical domain. We then discuss the potential applications of the proposed model and ideas on how to improve it even further.

Despite meriting the growing consensus between researchers and practitioners of ontology modeling, the Web Ontology Language OWL still has a modest presence in the communities of "traditional" web developers and software engineers. This resulted in hoarding the semantic web field in a rather small circle of people with a certain profile of expertise. In this paper we present OntoJIT, our novel approach toward a democratized semantic web where we bring OWL ontologies into the comfort-zone of end-application developers. We focus particularly on parsing OWL source files into executable ontologies in an object oriented programming paradigm. We finally demonstrate the dynamic code-base created as the result of parsing some reference OWL DL ontologies.

Using formal concept analysis, we propose a method for engineering ontology from MongoDB to effectively represent unstructured data. Our method consists of three main phases: (1) generating formal context from a MongoDB, (2) applying formal concept analysis to derive a concept lattice from that formal context, and (3) converting the obtained concept lattice to the first prototype of an ontology. We apply our method on NorthWind database and demonstrate how the proposed mapping rules can be used for learning an ontology from such database. At the end, we discuss about suggestions by which we can improve and generalize the method for more complex database examples.

Die graduelle Zugehörigkeit, das Kernkonzept der unscharfen Logik, kann in der Datenvisualisierung genutzt werden, um Zusammenhänge zwischen Merkmalen und Zielklassen darzustellen. Solche Zugehörigkeitsfunktionen können mit der IFC-Filter-Methode direkt aus den Daten extrahiert werden. Die Visualisierung dieser Zusammenhänge leuchtet ein, und präsentiert den Entscheidungsträgern eine leicht verständliche Darstellung. In Smart Energy-Systemen kann diese angewendet werden, um komplexe Vorhersagemodelle nachvollziehbar zu machen, und fördert somit die Datenintelligenz. An einem Fallbeispiel einer Datenanalyse in der Energiewirtschaft wird aufgezeigt, wie die Datenvisualisierung mit dem IFC-Filter in der Praxis aussehen kann. Dieses Kapitel stellt zudem den Bezug zu Big Data Management her, indem es die Datenvisualisierung für Smart Energy im größeren Zusammenhang eines Referenzmodells einordnet.

The extension of traditional data mining methods to time series has been effectively applied to a wide range of domains such as finance, econometrics, biology, security, and medicine. Many existing mining methods deal with the task of change points detection, but very few provide a flexible approach. Querying specific change points with linguistic variables is particularly useful in crime analysis, where intuitive, understandable, and appropriate detection of changes can significantly improve the allocation of resources for timely and concise operations. In this paper, we propose an on-line method for detecting and querying change points in crime-related time series with the use of a meaningful representation and a fuzzy inference system. Change points detection is based on a shape space representation, and linguistic terms describing geometric properties of the change points are used to express queries, offering the advantage of intuitiveness and flexibility. An empirical evaluation is first conducted on a crime data set to confirm the validity of the proposed method and then on a financial data set to test its general applicability. A comparison to a similar change-point detection algorithm and a sensitivity analysis are also conducted. Results show that the method is able to accurately detect change points at very low computational costs. More broadly, the detection of specific change points within time series of virtually any domain is made more intuitive and more understandable, even for experts not related to data mining.

In this paper we introduce PdsCART, a parallel decision tree learning algorithm. There are three characteristics that are important to emphasize and make this algorithm particularly interesting. Firstly, the algorithm we present here can work with streaming data, i.e. one pass over data is sufficient to construct the tree. Secondly, the algorithm is able to process in parallel a larger amount of data stream records and can therefor handle efficiently very large data sets. And thirdly, the algorithm can be implemented in the MapReduce framework. Details about the algorithm and some basic performance results are presented.

Weighted constraint satisfaction problems are difficult optimization problems that could model applications from various domains. Evolutionary algorithms are not the first option for solving such type of problems. In this work, the evolutionary algorithm uses the information extracted from the previous best solutions to guide the search in the next iterations. After the archive of previous best solutions has been sufficiently (re)filled, a data mining module is called to find association rules between variables and values. The generated rules are used to improve further the search process. Different methods of applying the association rules are investigated. Computational experiments are done on academic and real-world problem instances. The obtained results validate the approach and show that it is competitive with existing approaches in literature.

•We survey the literature on fuzzy classification.•We introduce a novel algorithm for membership function generation from data.•We present a software prototype that implements this algorithm.•We evaluate our system qualitatively based on two case studies in real organizations.•Conclusion: the resulting models can visually explain the data to human experts.

The evolutionary algorithms find difficulties in solving constraint satisfaction problems. The paper verifies if such algorithms could improve their results by using data mining techniques. The proposed approach uses association rules mining to guide the evolutionary search. The association rules are found from the past experience of the algorithm and are applied on individuals in order to keep the good direction and to improve them. A new escaping local optima strategy is proposed based on the mined rules. The considered problems to be solved are over-constrained constraint satisfaction problems where the number of satisfied constraints must be maximized. Results on randomly generated binary Max-CSP instances and on real world problems are given.

Die graduelle Zugehörigkeit, das Kernkonzept der unscharfen Logik, kann in der Datenanalyse genutzt werden, um Zusammenhänge darzustellen. Diese Grade der Zugehörigkeit können direkt aus den Daten extrahiert werden. Die Visualisierung dieser Zusammenhänge leuchtet ein und präsentiert den Entscheidungsträgern eine leicht verständliche Darstellung. An einem Fallbeispiel aus der Datenanalyse in der Energiewirtschaft wird aufgezeigt, wie diese Datenassoziierung in der Praxis aussehen kann. Der Beitrag stellt den Bezug zum Web Monitoring her, indem er eine entsprechende induktive und graduelle Konzeptassoziierung für die Analyse von Webdaten skizziert.

Grouping events having similarities has always been interesting for analysts. Actually, when a label is put on top of a set of events to denote they share common properties, the automation and the capability to conduct reasoning with this set drastically increase. This is particularly true when considering criminal events for crime analysts, conjunction, interpretation and explanation can be key success factors to apprehend criminals. In this paper, we present the CriLiM methodology for investigating both serious and high-volume crime. Our artifact consists in implementing a tailored computerized crime linkage system, based on a fuzzy MCDM approach in order to combine spatio-temporal, behavioral, and forensic information. As a proof of concept, series in burglaries are examined from real data and compared to expert results.