Digital news platforms rely on recommendation systems to meet the diverse information needs of readers. However, most research focuses on major, resource-rich languages, overlooking the linguistic diversity of online communities. Moreover, existing work typically assumes monolingual news consumption, neglecting polyglot users, and resulting in a lack of multilingual benchmarks for developing recommenders suited to multilingual and low-resource contexts. To address this gap, we introduce xMIND, an open, multilingual news recommendation dataset created by machine translating the English MIND dataset into 14 linguistically and geographically diverse languages with varying digital footprints. Using xMIND, we systematically evaluate several content-based neural news recommenders (NNRs) in zero-shot ( ZS-XLT ) and few-shot ( FS-XLT ) cross-lingual transfer, examining both monolingual and bilingual consumption patterns. In FS-XLT , we compare random and category-based replacement methods for incorporating target-language data during training. Our results show that (i) current NNRs, grounded in multilingual language models, experience significant performance drops in ZS-XLT , and (ii) injecting target-language data in FS-XLT provides limited improvements, especially for bilingual consumption. Notably, randomly injecting target-language news during training leads to greater performance gains compared to category-based replacements. Our in-depth analysis of representation alignment between source and target languages within the language model shows that FS-XLT improves cross-lingual alignment primarily for high-resource languages, while low-resource languages remain weakly aligned with English. These findings highlight the need for broader research efforts in multilingual and cross-lingual news recommendation. We release xMIND at https://github.com/andreeaiana/xMIND .

Knowledge graphs represent information as structured triples and serve as the backbone for a wide range of applications, including question answering, link prediction, and recommendation systems. A prominent line of research for exploring knowledge graphs involves graph embedding methods, where entities and relations are represented in low-dimensional vector spaces that capture underlying semantics and structure. However, most existing methods rely on assumptions such as the Closed World Assumption or Local Closed World Assumption, treating missing triples as false. This contrasts with the Open World Assumption underlying many real-world knowledge graphs. Furthermore, while explicitly stated negative statements can help distinguish between false and unknown triples, they are rarely included in knowledge graphs and are often overlooked during embedding training.

Business process models are widely used in enterprises to orchestrate workflows involving events, tasks, and agents. BPMN provides a standardized syntax to capture structural semantics, yet domain-level semantics often remain implicit, typically expressed through descriptive annotations. Thus, integrating BPMN processes with external data is challenging, limiting semantic interoperability. Building on the principles of Linked Data, several approaches have been proposed to embed BPMN processes within ontologies and knowledge graphs to overcome this limitation. However, current approaches are restricted to abstract process models, with limited or no support for dynamic runtime executions. Therefore, we introduce semantic BPMN (sBPMN), a framework for representing BPMN process models in a modular ontology directly derived from the official BPMN 2.0 XML schema, and their execution traces in a dynamic knowledge graph. Both static and dynamic fragments can then be easily linked to external sources. We demonstrate a system architecture compatible with state-of-the-art BPMN orchestration engines, showing that sBPMN enables semantic linking of process elements and external data as well as validation of process instances against formal constraints. Finally, we showcase the applicability of sBPMN in an industrial use case, where human–AI processes are modeled via BPMN and semantically enriched with domain data to facilitate standardization and semantic downstream analysis.

Knowledge Graphs are used in various domains to represent knowledge about entities and their relations.In the vast majority of cases, they capture what is known to be true about those entities, i.e., positive statements, while the Open World Assumption implicitly states that everything not expressed in the graph may or may not be true.This makes it difficult and less frequent to capture information explicitly known not to be true, i.e., negative statements.Moreover, while those negative statements could bear the potential to learn more useful representations in knowledge graph embeddings, that direction has been explored only rarely.However, in many domains, negative information is particularly interesting, for example, in recommender systems, where negative associations of users and items can help in learning better user representations, or in the biomedical domain, where the knowledge that a patient does exhibit a specific symptom can be crucial for accurate disease diagnosis.In this paper, we argue that negative statements should be given more attention in knowledge graph embeddings.Moreover, we investigate how they can be used in knowledge graph embedding methods, highlighting their potential in some interesting use cases.We discuss some existing works and preliminary results that incorporate explicitly declared negative statements in walk-based knowledge graph embedding methods.Finally, we outline promising avenues for future research in this area.

Knowledge graphs have become a universal data representation and integration mechanism. They recently gained interest in the spatial area. Spatial-Temporal Knowledge Graphs (STKGs) in particular have been created to integrate diverse sets of spatial data and model the relationships of spatial entities. Public knowledge graphs, such as KnowWhereGraph and WorldKG, have received a high traction in the domain of STKGs. In this paper, we compare three STKGs using downstream tasks within the Spatio-Temporal domain, and also discuss the underlying modeling decisions. We conduct an evaluation on a wildfire dataset and a housing dataset, comparing different embedding methodologies for the different knowledge graphs. We show that modeling paradigms in STKGs as well as algorithmic choices can have an impact on the downstream performance, and discuss challenges in both areas.

Digital news platforms use news recommenders as the main instrument to cater to the individual information needs of readers. Despite an increasingly language-diverse online community, in which many Internet users consume news in multiple languages, the majority of news recommendation focuses on major, resource-rich languages. Moreover, nearly all news recommendation efforts assume monolingual news consumption, whereas more and more users tend to consume information in at least two languages. Accordingly, the existing body of work on news recommendation suffers from a lack of publicly available multilingual benchmarks that would catalyze development of news recommenders effective in multilingual settings and for low-resource languages. Aiming to fill this gap, we introduce xMIND, an open, multilingual news recommendation dataset derived from the English MIND dataset using machine translation, covering a set of 14 linguistically and geographically diverse languages, with digital footprints of varying sizes. Using xMIND, we systematically benchmark several content-based neural news recommenders (NNRs) in zero-shot (ZS-XLT) and few-shot (FS-XLT) cross-lingual transfer scenarios, considering both monolingual and bilingual news consumption patterns. Our findings reveal that (i) current NNRs, even when based on a multilingual language model, suffer from substantial performance losses under ZS-XLT and that (ii) inclusion of target-language data in FS-XLT training has limited benefits, particularly when combined with a bilingual news consumption. Our findings thus warrant a broader research effort in multilingual and cross-lingual news recommendation. We release xMIND at https://github.com/andreeaiana/xMIND.

In recent years, countless research papers have addressed the topics of knowledge graph creation, extension, or completion in order to create knowledge graphs that are larger, more correct, or more diverse. This research is typically motivated by the argumentation that using such enhanced knowledge graphs to solve downstream tasks will improve performance. Nonetheless, this is hardly ever evaluated. Instead, the predominant evaluation metrics - aiming at correctness and completeness - are undoubtedly valuable but fail to capture the complete picture, i.e., how useful the created or enhanced knowledge graph actually is. Further, the accessibility of such a knowledge graph is rarely considered (e.g., whether it contains expressive labels, descriptions, and sufficient context information to link textual mentions to the entities of the knowledge graph). To better judge how well knowledge graphs perform on actual tasks, we present KGrEaT - a framework to estimate the quality of knowledge graphs via actual downstream tasks like classification, clustering, or recommendation. Instead of comparing different methods of processing knowledge graphs with respect to a single task, the purpose of KGrEaT is to compare various knowledge graphs as such by evaluating them on a fixed task setup. The framework takes a knowledge graph as input, automatically maps it to the datasets to be evaluated on, and computes performance metrics for the defined tasks. It is built in a modular way to be easily extendable with additional tasks and datasets.

Ontology (and more generally: Knowledge Graph) Matching is a challenging task where information in natural language is one of the most important signals to process. With the rise of Large Language Models, it is possible to incorporate this knowledge in a better way into the matching pipeline. A number of decisions still need to be taken, e.g., how to generate a prompt that is useful to the model, how information in the KG can be formulated in prompts, which Large Language Model to choose, how to provide existing correspondences to the model, how to generate candidates, etc. In this paper, we present a prototype that explores these questions by applying zero-shot and few-shot prompting with multiple open Large Language Models to different tasks of the Ontology Alignment Evaluation Initiative (OAEI). We show that with only a handful of examples and a well-designed prompt, it is possible to achieve results that are en par with supervised matching systems which use a much larger portion of the ground truth.

Knowledge graph embedding models (KGEMs) have gained considerable traction in recent years. These models learn a vector representation of knowledge graph entities and relations, a.k.a. knowledge graph embeddings (KGEs). Learning versatile KGEs is desirable as it makes them useful for a broad range of tasks. However, KGEMs are usually trained for a specific task, which makes their embeddings task-dependent. In parallel, the widespread assumption that KGEMs actually create a semantic representation of the underlying entities and relations (e.g., project similar entities closer than dissimilar ones) has been challenged. In this work, we design heuristics for generating protographs – small, modified versions of a KG that leverage RDF/S information. The learnt protograph-based embeddings are meant to encapsulate the semantics of a KG, and can be leveraged in learning KGEs that, in turn, also better capture semantics. Extensive experiments on various evaluation benchmarks demonstrate the soundness of this approach, which we call Modular and Agnostic SCHema-based Integration of protograph Embeddings (MASCHInE). In particular, MASCHInE helps produce more versatile KGEs that yield substantially better performance for entity clustering and node classification tasks. For link prediction, using MASCHinE substantially increases the number of semantically valid predictions with equivalent rank-based performance.

The advent of personalized news recommendation has given rise to increasingly complex recommender architectures. Most neural news recommenders rely on user click behavior and typically introduce dedicated user encoders that aggregate the content of clicked news into user embeddings (early fusion). These models are predominantly trained with standard point-wise classification objectives. The existing body of work exhibits two main shortcomings: (1) despite general design homogeneity, direct comparisons between models are hindered by varying evaluation datasets and protocols; (2) it leaves alternative model designs and training objectives vastly unexplored. In this work, we present a unified framework for news recommendation, allowing for a systematic and fair comparison of news recommenders across several crucial design dimensions: (i) candidate-awareness in user modeling, (ii) click behavior fusion, and (iii) training objectives. Our findings challenge the status quo in neural news recommendation. We show that replacing sizable user encoders with parameter-efficient dot products between candidate and clicked news embeddings (late fusion) often yields substantial performance gains. Moreover, our results render contrastive training a viable alternative to point-wise classification objectives.

The Semantic Web is distributed yet interoperable: Distributed since resources are created and published by a variety of producers, tailored to their specific needs and knowledge; Interoperable as entities are linked across resources, allowing to use resources from different providers in concord. Complementary to the explicit usage of Semantic Web resources, embedding methods made them applicable to machine learning tasks. Subsequently, embedding models for numerous tasks and structures have been developed, and embedding spaces for various resources have been published. The ecosystem of embedding spaces is distributed but not interoperable: Entity embeddings are not readily comparable across different spaces. To parallel the Web of Data with a Web of Embeddings, we must thus integrate available embedding spaces into a uniform space. Current integration approaches are limited to two spaces and presume that both of them were embedded with the same method — both assumptions are unlikely to hold in the context of a Web of Embeddings. In this paper, we present FedCoder— an approach that integrates multiple embedding spaces via a latent space. We assert that linked entities have a similar representation in the latent space so that entities become comparable across embedding spaces. FedCoder employs an autoencoder to learn this latent space from linked as well as non-linked entities. Our experiments show that FedCoder substantially outperforms state-of-the-art approaches when faced with different embedding models, that it scales better than previous methods in the number of embedding spaces, and that it improves with more graphs being integrated whilst performing comparably with current approaches that assumed joint learning of the embeddings and were, usually, limited to two sources. Our results demonstrate that FedCoder is well adapted to integrate the distributed, diverse, and large ecosystem of embeddings spaces into an interoperable Web of Embeddings.

Despite the vast amount of information encoded in knowledge graphs, they often remain incomplete. Neural networks, in particular Graph Convolutional Neural Networks, have been shown to be effective predictors to complete information about the class affiliation of entities in knowledge graphs. However, these models remain ignorant to their predictions confidence due to their used point estimate of a softmax output. In this paper, we combine Graph Convolutional Neural Networks with recent developments in the field of Evidential Learning by placing a Dirichlet distribution on the class probabilities to overcome this problem. We use the continuous output of a Graph Convolutional Neural Network as parameters for a Dirichlet distribution. In this way, the predictions of the model are represented as a distribution over possible softmax outputs, rather than a point estimate of a softmax output. The experiments show that a better performance in predicting class affiliations can be achieved compared to recent models. In addition, the experiments show that this approach overcomes the well-known problem of overconfident prediction of deterministic neural networks.

Knowledge graphs (KGs) provide information in machine interpretable form. In cases where multiple KGs are used in the same system, that information needs to be integrated. This is usually done by automated matching systems. Most of those systems consider only 1:1 (binary) matching tasks. Thus, matching a larger number of knowledge graphs with such systems would lead to quadratic efforts. In this paper, we empirically analyze different approaches to reduce the task of multi-source matching to a linear number of executions of binary matching systems. We show that the matching order of KGs and the multi-source strategy actually matter and that near-optimal results can be achieved with linear efforts.

Public knowledge graphs such as DBpedia and Wikidata have been recognized as interesting sources of background knowledge to build content-based recommender systems. They can be used to add information about the items to be recommended and links between those. While quite a few approaches for exploiting knowledge graphs have been proposed, most of them aim at optimizing the recommendation strategy while using a fixed knowledge graph. In this paper, we take a different approach, i.e., we fix the recommendation strategy and observe changes when using different underlying knowledge graphs. Particularly, we use different language editions of DBpedia. We show that the usage of different knowledge graphs does not only lead to differently biased recommender systems, but also to recommender systems that differ in performance for particular fields of recommendations.

Knowledge about entities and their interrelations is a crucial factor of success for tasks like question answering or text summarization. Publicly available knowledge graphs like Wikidata or DBpedia are, however, far from being complete. In this paper, we explore how information extracted from similar entities that co-occur in structures like tables or lists can help to increase the coverage of such knowledge graphs. In contrast to existing approaches, we do not focus on relationships within a listing (e.g., between two entities in a table row) but on the relationship between a listing’s subject entities and the context of the listing. To that end, we propose a descriptive rule mining approach that uses distant supervision to derive rules for these relationships based on a listing’s context. Extracted from a suitable data corpus, the rules can be used to extend a knowledge graph with novel entities and assertions. In our experiments we demonstrate that the approach is able to extract up to 3M novel entities and 30M additional assertions from listings in Wikipedia. We find that the extracted information is of high quality and thus suitable to extend Wikipedia-based knowledge graphs like DBpedia, YAGO, and CaLiGraph. For the case of DBpedia, this would result in an increase of covered entities by roughly 50%.

There has been a rapid growth in the use of symbolic representations along with their applications in many important tasks. Symbolic representations, in the form of Knowledge Graphs (KGs), constitute large networks of real-world entities and their relationships. On the other hand, sub-symbolic artificial intelligence has also become a mainstream area of research. This workshop brought together researchers to discuss and foster collaborations on the intersection of these two areas.

Popular cross-domain knowledge graphs, such as DBpedia and YAGO, are built from Wikipedia, and therefore similar in coverage. In contrast, Wikifarms like Fandom contain Wikis for specific topics, which are often complementary to the information contained in Wikipedia, and thus DBpedia and YAGO. Extracting these Wikis with the DBpedia extraction framework is possible, but results in many isolated knowledge graphs. In this paper, we show how to create one consolidated knowledge graph, called DBkWik, from thousands of Wikis. We perform entity resolution and schema matching, and show that the resulting large-scale knowledge graph is complementary to DBpedia. Furthermore, we discuss the potential use of DBkWik as a benchmark for knowledge graph matching.

The problem of automatic detection of fake news in social media, e.g., on Twitter, has recently drawn some attention. Although, from a technical perspective, it can be regarded as a straight-forward, binary classification problem, the major challenge is the collection of large enough training corpora, since manual annotation of tweets as fake or non-fake news is an expensive and tedious endeavor. In this paper, we discuss a weakly supervised approach, which automatically collects a large-scale, but very noisy training dataset comprising hundreds of thousands of tweets. During collection, we automatically label tweets by their source, i.e., trustworthy or untrustworthy source, and train a classifier on this dataset. We then use that classifier for a different classification target, i.e., the classification of fake and non-fake tweets. Although the labels are not accurate according to the new classification target (not all tweets by an untrustworthy source need to be fake news, and vice versa), we show that despite this unclean inaccurate dataset, it is possible to detect fake news with an F1 score of up to 0.9.

Entity matching, or record linkage, is the task of identifying records that refer to the same entity.Naive entity matching techniques (i.e., brute-force pairwise comparisons) have quadratic complexity.A typical shortcut to the problem is to employ blocking techniques to reduce the number of comparisons, i.e. to partition the data in several blocks and only compare records within the same block.While classic blocking methods are designed for data from relational databases with clearly defined schemas, they are not applicable to data from Web tables, which are more prone to noise and do not come with an explicit schema.At the same time, Web tables are an interesting data source for many knowledge intensive tasks, which makes record linkage on Web Tables an important challenge.In this work, we propose an unsupervised approach to partition the data, that does not exploit any external knowledge, but only relies on heuristics to select the blocking attributes.We compare different partitioning methods: we use (i) clustering on bagof-words, (ii) binning via Locality-Sensitive Hashing and (iii) clustering using word embeddings.In particular, the clustering methods show good results on a standard dataset of Web Tables, and, when combined with word embeddings, are a robust solution which allows for computing the clusters in a dense, low-dimensional space.

Although the link prediction problem, where missing relation assertions are predicted, has been widely researched, error detection did not receive as much attention. In this paper, we investigate the problem of error detection in relation assertions of knowledge graphs, and we propose an error detection method which relies on path and type features used by a classifier for every relation in the graph exploiting local feature selection. We perform an extensive evaluation on a variety of datasets, backed by a manual evaluation on DBpedia and NELL, and we propose and evaluate heuristics for the selection of relevant graph paths to be used as features in our method.

Taxonomies are an important ingredient of knowledge organization, and serve as a backbone for more sophisticated knowledge representations in intelligent systems, such as formal ontologies. However, building taxonomies manually is a costly endeavor, and hence, automatic methods for taxonomy induction are a good alternative to build large-scale taxonomies. In this paper, we propose TIEmb, an approach for automatic unsupervised class subsumption axiom extraction from knowledge bases using entity and text embeddings. We apply the approach on the WebIsA database, a database of subsumption relations extracted from the large portion of the World Wide Web, to extract class hierarchies in the Person and Place domain.

Linked Data on the Web is either created from structured data sources (such as relational databases), from semi-structured sources (such as Wikipedia), or from unstructured sources (such as text). In the latter two cases, the generated Linked Data will likely be noisy and incomplete. In this paper, we present two algorithms that exploit statistical distributions of properties and types for enhancing the quality of incomplete and noisy Linked Data sets: SDType adds missing type statements, and SDValidate identifies faulty statements. Neither of the algorithms uses external knowledge, i.e., they operate only on the data itself. We evaluate the algorithms on the DBpedia and NELL knowledge bases, showing that they are both accurate as well as scalable. Both algorithms have been used for building the DBpedia 3.9 release: With SDType, 3.4 million missing type statements have been added, while using SDValidate, 13,000 erroneous RDF statements have been removed from the knowledge base.

Ontologies have been increasingly used in software systems in the past years. However, in many of those systems, the ontologies are hidden “under the hood”. While a lot of useful applications of ontologies on the database and business logic layer have been proposed, the employment of ontologies in user interfaces has been gaining comparatively little attention so far. For providing a deeper understanding of that field as well as assisting developers of ontology-enhanced user interfaces, the authors give an overview of such applications and introduce a schema for characterizing the requirements of ontology-enhanced user interfaces. With this article, a state of the art survey of approaches is presented along with promising research directions.