Axel-Cyrille Ngonga Ngomo’s research while affiliated with Paderborn University and other places

Clifford algebras are a natural extension of division algebras, including real numbers, complex numbers, quaternions, and octonions. Previous research in knowledge graph embeddings has focused exclusively on Clifford algebras of a specific type, which do not include nilpotent base vectors—elements that square to zero. In this work, we introduce a novel approach by incorporating nilpotent base vectors with a nilpotency index of two, leading to a more general form of Clifford algebras named degenerate Clifford algebras. This generalization to degenerate Clifford algebras does allow for covering dual numbers and as such include translations and rotations models under the same generalization paradigm for the first time. We develop two models to determine the parameters that define the algebra: one using a greedy search and another predicting the parameters based on neural network embeddings of the input knowledge graph. Our evaluation on seven benchmark datasets demonstrates that this incorporation of nilpotent vectors enhances the quality of embeddings. Additionally, our method outperforms state-of-the-art approaches in terms of generalization, particularly regarding the mean reciprocal rank achieved on validation data. Finally, we show that even a simple greedy search can effectively discover optimal or near-optimal parameters for the algebra.

Knowledge Graph Embedding (KGE) transforms a discrete Knowledge Graph (KG) into a continuous vector space facilitating its use in various AI-driven applications like Semantic Search, Question Answering, or Recommenders. While KGE approaches are effective in these applications, most existing approaches assume that all information in the given KG is correct. This enables attackers to influence the output of these approaches, e.g., by perturbing the input. Consequently, the robustness of such KGE approaches has to be addressed. Recent work focused on adversarial attacks. However, non-adversarial attacks on all attack surfaces of these approaches have not been thoroughly examined. We close this gap by evaluating the impact of non-adversarial attacks on the performance of 5 state-of-the-art KGE algorithms on 5 datasets with respect to attacks on 3 attack surfaces-graph, parameter, and label perturbation. Our evaluation results suggest that label perturbation has a strong effect on the KGE performance, followed by parameter perturbation with a moderate and graph with a low effect.

We discuss interpretability and explainability of machine learning models. We introduce a universal interpretability index, JJ, to quantify and monitor the interpretability of a general-purpose model, which can be static or evolve incre-mentally from a data stream. The models can be transparent classifiers, predictors or controllers operating on partitions or granules of the data space, e.g., rule-based models, trees, proba-bilistic clustering models, modular or granular neural networks. Additionally, black boxes can be monitored either after the derivation of a global or local surrogate model as a result of the application of a model-agnostic explainer. The index does not de-pend on the type of algorithm that creates or updates the model, i.e., supervised, unsupervised, semi-supervised, or reinforcement. While loss or error-related indices, validity measures, processing time, and closed-loop criteria have been used to evaluate model effectiveness across different fields, a general interpretability index in consonance with explainable AI does not exist. The index JJ is computed straightforwardly. It reflects the principle of justifiable granularity by taking into account balanced volumes, number of partitions and dependent parameters, and features per partition. The index advocates that a concise model founded on balanced partitions offers a higher level of interpretability. It facilitates comparisons between models, can be used as a term in loss functions or embedded into learning procedures to motivate equilibrium of volumes, and ultimately support human-centered decision making. Computational experiments show the versatility of the index in a biomedical prediction problem from speech data, and in image classification.

We introduce a modified incremental learning algorithm for evolving Granular Neural Network Classifiers (eGNN-C+). We use double-boundary hyper-boxes to represent granules, and customize the adaptation procedures to enhance the robustness of outer boxes for data coverage and noise suppression, while ensuring that inner boxes remain flexible to capture drifts. The classifier evolves from scratch, incorporates new classes on the fly, and performs local incremental feature weighting. As an application, we focus on the classification of emotion-related patterns within electroencephalogram (EEG) signals. Emotion recognition is crucial for enhancing the realism and interactivity of computer systems. The challenge lies exactly in developing high-performance algorithms capable of effectively managing individual differences and non-stationarities in physiological data without relying on subject-specific information. We extract features from the Fourier spectrum of EEG signals obtained from 28 individuals engaged in playing computer games – a public dataset. Each game elicits a different predominant emotion: boredom, calmness, horror, or joy. We analyze individual electrodes, time window lengths, and frequency bands to assess the accuracy and interpretability of resulting user-independent neural models. The findings indicate that both brain hemispheres assist classification, especially electrodes on the temporal (T8) and parietal (P7) areas, alongside contributions from frontal and occipital electrodes. While patterns may manifest in any band, the Alpha (8-13Hz), Delta (1-4Hz), and Theta (4-8Hz) bands, in this order, exhibited higher correspondence with the emotion classes. The eGNN-C+ demonstrates effectiveness in learning EEG data. It achieves an accuracy of 81.7% and a 0.0029 II interpretability using 10-second time windows, even in face of a highly-stochastic time-varying 4-class classification problem.

We discover a connection between finding subset-maximal repairs for sets of functional and inclusion dependencies, and computing extensions within argumentation frameworks (AFs). We study the complexity of existence of a repair, and deciding whether a given tuple belongs to some (or every) repair, by simulating the instances of these problems via AFs. We prove that subset-maximal repairs under functional dependencies correspond to the naive extensions, which also coincide with the preferred and stable extensions in the resulting AFs. For inclusion dependencies one needs a pre-processing step on the resulting AFs in order for the extensions to coincide. Allowing both types of dependencies breaks this relationship between extensions and only preferred semantics captures the repairs. Finally, we establish that the complexities of the above decision problems are ${\textbf {NP}}$-complete and $\boldsymbol{\mathrm {\Pi }}^{ {\textbf {P}}}_2$-complete, when both functional and inclusion dependencies are allowed.

We introduce a modified incremental learning algorithm for evolving Granular Neural Network Classifiers (eGNN-C+). We use double-boundary hyper-boxes to represent granules, and customize the adaptation procedures to enhance the robust-ness of outer boxes for data coverage and noise suppression, while ensuring that inner boxes remain flexible to capture drifts. The classifier evolves from scratch, incorporates new classes on the fly, and performs local incremental feature weighting. As an application, we focus on the classification of emotion-related patterns within electroencephalogram (EEG) signals. Emotion recognition is crucial for enhancing the realism and interactivity of computer systems. The challenge lies exactly in developing high-performance algorithms capable of effectively managing individual differences and non-stationarities in physiological data without relying on subject-specific calibration data. We extract features from the Fourier spectrum of EEG signals obtained from 28 individuals engaged in playing computer games-a public dataset. Each game elicits a different predominant emotion: boredom, calmness, horror, or joy. We analyze individual electrodes, time window lengths, and frequency bands to assess the accuracy and interpretability of resulting user-independent neural models. The findings indicate that both brain hemispheres assist classification, especially electrodes on the temporal (T8) and parietal (P7) areas, alongside contributions from frontal and occipital electrodes. While patterns may manifest in any band, the Alpha (8-13Hz), Delta (1-4Hz), and Theta (4-8Hz) bands, in this order, exhibited higher correspondence with the emotion classes. The eGNN-C+ demonstrates effectiveness in learning EEG data. It achieves an accuracy of 81.7% and a 0.0029 II interpretability using 10-second time windows, even in face of a highly-stochastic time-varying 4-class classification problem.

Knowledge bases are now first-class citizens of the Web. Circa 50% of the 3.2 billion websites in the 2022 crawl of Web Data Commons contains knowledge base fragments in RDF. The 82 billion assertions known to exist in these websites are complemented by a roughly comparable number of triples available in dumps. As this data is now the backbone of a number of applications, it stands to reason that machine learning approaches able to exploit the explicit semantics exposed by RDF knowledge bases must scale to large knowledge bases. In this chapter, we present approaches based on continuous and symbolic representations that aim to achieve this goal by addressing some of the main scalability bottlenecks of existing class expression learning approaches. While we focus on the description logic ALC , the approaches we present are far from being limited to this particular expressiveness.

The Nomenclature of Territorial Units for Statistics (NUTS) is a classification that represents countries in the European Union (EU). It is published at intervals of several years and organized in a hierarchical system where geographical areas are subdivided according to their population sizes. In addition to NUTS, there is a further subdivided hierarchy level, named Local Administrative Units (LAU), whose data are updated annually by EU member states. While both datasets are published by Eurostat as Excel files, an additional RDF dataset is available for NUTS up to the 2016 scheme. With this work, we provide the Linked Data community with an up-to-date Knowledge Graph in which NUTS and LAU data are linked and which contains population numbers as well as area sizes. We also publish an Open Source generator software for future released versions that will naturally arise due to changes in population numbers. These contributions can be used to enrich other datasets and allow comparisons among regions in the European Union. All resources are available at https://w3id.org/launuts.KeywordsEUEuropean UnionEurostatKnowledge GraphLAULauNutsLinked DataNUTS

Relation extraction plays an important role in natural language processing. There is a wide range of available datasets that benchmark existing relation extraction approaches. However, most benchmarking datasets are provided in different formats containing specific annotation rules, thus making it difficult to conduct experiments on different types of relation extraction approaches. We present RELD, an RDF knowledge graph of eight open-licensed and publicly available relation extraction datasets. We modeled the benchmarking datasets into a single ontology that provides a unified format for data access, along with annotations required for training different types of relation extraction systems. Moreover, RELD abides by the Linked Data principles. To the best of our knowledge, RELD is the largest RDF knowledge graph of entities and relations from text, containing $\sim$1230 million triples describing 1034 relations, 2 million sentences, 3 million abstracts and 4013 documents. RELD contributes to a variety of uses in the natural language processing community, and distinctly provides unified and easy modeling of data for benchmarking relation extraction and named entity recognition models.KeywordsKnowledge graphRelation extractionbenchmarksNatural language processing.ontologyRDF