Nada Lavrač

Jan Kralj, Blaž Škrlj, Živa Ramšak et al.

Biological systems can be studied at multiple levels of information, including gene, protein, RNA and different interaction networks levels. The goal of this work is to explore how the fusion of systems' level information with temporal dynamics of gene expression can be used in combination with non-linear approximation power of deep neural networks to predict novel gene functions in a non-model organism potato \emph{Solanum tuberosum}. We propose DDeMON (Dynamic Deep learning from temporal Multiplex Ontology-annotated Networks), an approach for scalable, systems-level inference of function annotation using time-dependent multiscale biological information. The proposed method, which is capable of considering billions of potential links between the genes of interest, was applied on experimental gene expression data and the background knowledge network to reliably classify genes with unknown function into five different functional ontology categories, linked to the experimental data set. Predicted novel functions of genes were validated using extensive protein domain search approach.

Luka Andrenšek, Boshko Koloski, Andraž Pelicon et al.

We investigate zero-shot cross-lingual news sentiment detection, aiming to develop robust sentiment classifiers that can be deployed across multiple languages without target-language training data. We introduce novel evaluation datasets in several less-resourced languages, and experiment with a range of approaches including the use of machine translation; in-context learning with large language models; and various intermediate training regimes including a novel task objective, POA, that leverages paragraph-level information. Our results demonstrate significant improvements over the state of the art, with in-context learning generally giving the best performance, but with the novel POA approach giving a competitive alternative with much lower computational overhead. We also show that language similarity is not in itself sufficient for predicting the success of cross-lingual transfer, but that similarity in semantic content and structure can be equally important.

Boshko Koloski, Nada Lavrač, Senja Pollak et al.

In the domain of semi-supervised learning, the current approaches insufficiently exploit the potential of considering inter-instance relationships among (un)labeled data. In this work, we address this limitation by providing an approach for inferring latent graphs that capture the intrinsic data relationships. By leveraging graph-based representations, our approach facilitates the seamless propagation of information throughout the graph, effectively incorporating global and local knowledge. Through evaluations on biomedical tabular datasets, we compare the capabilities of our approach to other contemporary methods. Our work demonstrates the significance of inter-instance relationship discovery as practical means for constructing robust latent graphs to enhance semi-supervised learning techniques. The experiments show that the proposed methodology outperforms contemporary state-of-the-art methods for (semi-)supervised learning on three biomedical datasets.

Marko Pranjić, Boshko Koloski, Nada Lavrač et al.

Neighborhood graphs are a critical but often fragile step in spectral clustering of text embeddings. On realistic text datasets, standard $k$-NN graphs can contain many disconnected components at practical sparsity levels (small $k$), making spectral clustering degenerate and sensitive to hyperparameters. We introduce a simple incremental $k$-NN graph construction that preserves connectivity by design: each new node is linked to its $k$ nearest previously inserted nodes, which guarantees a connected graph for any $k$. We provide an inductive proof of connectedness and discuss implications for incremental updates when new documents arrive. We validate the approach on spectral clustering of SentenceTransformer embeddings using Laplacian eigenmaps across six clustering datasets from the Massive Text Embedding Benchmark. Compared to standard $k$-NN graphs, our method outperforms in the low-$k$ regime where disconnected components are prevalent, and matches standard $k$-NN at larger $k$.

Boshko Koloski, Nada Lavrač, Bojan Cestnik et al.

In an era marked by a rapid increase in scientific publications, researchers grapple with the challenge of keeping pace with field-specific advances. We present the `AHAM' methodology and a metric that guides the domain-specific \textbf{adapt}ation of the BERTopic topic modeling framework to improve scientific text analysis. By utilizing the LLaMa2 generative language model, we generate topic definitions via one-shot learning by crafting prompts with the \textbf{help} of domain experts to guide the LLM for literature mining by \textbf{asking} it to model the topic names. For inter-topic similarity evaluation, we leverage metrics from language generation and translation processes to assess lexical and semantic similarity of the generated topics. Our system aims to reduce both the ratio of outlier topics to the total number of topics and the similarity between topic definitions. The methodology has been assessed on a newly gathered corpus of scientific papers on literature-based discovery. Through rigorous evaluation by domain experts, AHAM has been validated as effective in uncovering intriguing and novel insights within broad research areas. We explore the impact of domain adaptation of sentence-transformers for the task of topic \textbf{model}ing using two datasets, each specialized to specific scientific domains within arXiv and medarxiv. We evaluate the impact of data size, the niche of adaptation, and the importance of domain adaptation. Our results suggest a strong interaction between domain adaptation and topic modeling precision in terms of outliers and topic definitions.

Blaž Škrlj, Boshko Koloski, Senja Pollak et al.

Integrating structured knowledge from Knowledge Graphs (KGs) into Large Language Models (LLMs) enhances factual grounding and reasoning capabilities. This survey paper systematically examines the synergy between KGs and LLMs, categorizing existing approaches into two main groups: KG-enhanced LLMs, which improve reasoning, reduce hallucinations, and enable complex question answering; and LLM-augmented KGs, which facilitate KG construction, completion, and querying. Through comprehensive analysis, we identify critical gaps and highlight the mutual benefits of structured knowledge integration. Compared to existing surveys, our study uniquely emphasizes scalability, computational efficiency, and data quality. Finally, we propose future research directions, including neuro-symbolic integration, dynamic KG updating, data reliability, and ethical considerations, paving the way for intelligent systems capable of managing more complex real-world knowledge tasks.

Andrej Kastrin, Bojan Cestnik, Nada Lavrač

The explosive growth of scientific publications has created an urgent need for automated methods that facilitate knowledge synthesis and hypothesis generation. Literature-based discovery (LBD) addresses this challenge by uncovering previously unknown associations between disparate domains. This article surveys recent methodological advances in LBD, focusing on developments from 2000 to the present. We review progress in three key areas: knowledge graph construction, deep learning approaches, and the integration of pre-trained and large language models (LLMs). While LBD has made notable progress, several fundamental challenges remain unresolved, particularly concerning scalability, reliance on structured data, and the need for extensive manual curation. By examining ongoing advances and outlining promising future directions, this survey underscores the transformative role of LLMs in enhancing LBD and aims to support researchers and practitioners in harnessing these technologies to accelerate scientific innovation.

Andraž Repar, Nada Lavrač, Senja Pollak

Automated terminology extraction refers to the task of extracting meaningful terms from domain-specific texts. This paper proposes a novel machine learning approach to terminology extraction, which combines features from traditional term extraction systems with novel contextual features derived from contextual word embeddings. Instead of using a predefined list of part-of-speech patterns, we first analyse a new term-annotated corpus RSDO5 for the Slovenian language and devise a set of rules for term candidate selection and then generate statistical, linguistic and context-based features. We use a support-vector machine algorithm to train a classification model, evaluate it on the four domains (biomechanics, linguistics, chemistry, veterinary) of the RSDO5 corpus and compare the results with state-of-art term extraction approaches for the Slovenian language. Our approach provides significant improvements in terms of F1 score over the previous state-of-the-art, which proves that contextual word embeddings are valuable for improving term extraction.

Bojan Cestnik, Andrej Kastrin, Boshko Koloski et al.

By connecting disparate sources of scientific literature, literature\-/based discovery (LBD) methods help to uncover new knowledge and generate new research hypotheses that cannot be found from domain-specific documents alone. Our work focuses on bisociative LBD methods that combine bisociative reasoning with LBD techniques. The paper presents LBD through the lens of reproducible science to ensure the reproducibility of LBD experiments, overcome the inconsistent use of benchmark datasets and methods, trigger collaboration, and advance the LBD field toward more robust and impactful scientific discoveries. The main novelty of this study is a collection of Jupyter Notebooks that illustrate the steps of the bisociative LBD process, including data acquisition, text preprocessing, hypothesis formulation, and evaluation. The contributed notebooks implement a selection of traditional LBD approaches, as well as our own ensemble-based, outlier-based, and link prediction-based approaches. The reader can benefit from hands-on experience with LBD through open access to benchmark datasets, code reuse, and a ready-to-run Docker recipe that ensures reproducibility of the selected LBD methods.

Boshko Koloski, Nada Lavrač, Blaž Škrlj

Construction of neural network architectures suitable for learning from both continuous and discrete tabular data is a challenging research endeavor. Contemporary high-dimensional tabular data sets are often characterized by a relatively small instance count, requiring data-efficient learning. We propose HorNets (Horn Networks), a neural network architecture with state-of-the-art performance on synthetic and real-life data sets from scarce-data tabular domains. HorNets are based on a clipped polynomial-like activation function, extended by a custom discrete-continuous routing mechanism that decides which part of the neural network to optimize based on the input's cardinality. By explicitly modeling parts of the feature combination space or combining whole space in a linear attention-like manner, HorNets dynamically decide which mode of operation is the most suitable for a given piece of data with no explicit supervision. This architecture is one of the few approaches that reliably retrieves logical clauses (including noisy XNOR) and achieves state-of-the-art classification performance on 14 real-life biomedical high-dimensional data sets. HorNets are made freely available under a permissive license alongside a synthetic generator of categorical benchmarks.

Sebastian Mežnar, Matej Bevec, Nada Lavrač et al.

The increasing amounts of semantic resources offer valuable storage of human knowledge; however, the probability of wrong entries increases with the increased size. The development of approaches that identify potentially spurious parts of a given knowledge base is thus becoming an increasingly important area of interest. In this work, we present a systematic evaluation of whether structure-only link analysis methods can already offer a scalable means to detecting possible anomalies, as well as potentially interesting novel relation candidates. Evaluating thirteen methods on eight different semantic resources, including Gene Ontology, Food Ontology, Marine Ontology and similar, we demonstrated that structure-only link analysis could offer scalable anomaly detection for a subset of the data sets. Further, we demonstrated that by considering symbolic node embedding, explanations of the predictions (links) could be obtained, making this branch of methods potentially more valuable than the black-box only ones. To our knowledge, this is currently one of the most extensive systematic studies of the applicability of different types of link analysis methods across semantic resources from different domains.

Blaž Škrlj, Marko Jukič, Nika Eržen et al.

The COVID-19 pandemic triggered a wave of novel scientific literature that is impossible to inspect and study in a reasonable time frame manually. Current machine learning methods offer to project such body of literature into the vector space, where similar documents are located close to each other, offering an insightful exploration of scientific papers and other knowledge sources associated with COVID-19. However, to start searching, such texts need to be appropriately annotated, which is seldom the case due to the lack of human resources. In our system, the current body of COVID-19-related literature is annotated using unsupervised keyphrase extraction, facilitating the initial queries to the latent space containing the learned document embeddings (low-dimensional representations). The solution is accessible through a web server capable of interactive search, term ranking, and exploration of potentially interesting literature. We demonstrate the usefulness of the approach via case studies from the medicinal chemistry domain.

Timen Stepišnik Perdih, Nada Lavrač, Blaž Škrlj

With the wide adoption of black-box models, instance-based \emph{post hoc} explanation tools, such as LIME and SHAP became increasingly popular. These tools produce explanations, pinpointing contributions of key features associated with a given prediction. However, the obtained explanations remain at the raw feature level and are not necessarily understandable by a human expert without extensive domain knowledge. We propose ReEx (Reasoning with Explanations), a method applicable to explanations generated by arbitrary instance-level explainers, such as SHAP. By using background knowledge in the form of ontologies, ReEx generalizes instance explanations in a least general generalization-like manner. The resulting symbolic descriptions are specific for individual classes and offer generalizations based on the explainer's output. The derived semantic explanations are potentially more informative, as they describe the key attributes in the context of more general background knowledge, e.g., at the biological process level. We showcase ReEx's performance on nine biological data sets, showing that compact, semantic explanations can be obtained and are more informative than generic ontology mappings that link terms directly to feature names. ReEx is offered as a simple-to-use Python library and is compatible with tools such as SHAP and similar. To our knowledge, this is one of the first methods that directly couples semantic reasoning with contemporary model explanation methods. This paper is a preprint. Full version's doi is: 10.1109/SAMI50585.2021.9378668

Sebastian Mežnar, Nada Lavrač, Blaž Škrlj

Understanding how information propagates in real-life complex networks yields a better understanding of dynamic processes such as misinformation or epidemic spreading. The recently introduced branch of machine learning methods for learning node representations offers many novel applications, one of them being the task of spreading prediction addressed in this paper. We explore the utility of the state-of-the-art node representation learners when used to assess the effects of spreading from a given node, estimated via extensive simulations. Further, as many real-life networks are topologically similar, we systematically investigate whether the learned models generalize to previously unseen networks, showing that in some cases very good model transfer can be obtained. This work is one of the first to explore transferability of the learned representations for the task of node regression; we show there exist pairs of networks with similar structure between which the trained models can be transferred (zero-shot), and demonstrate their competitive performance. To our knowledge, this is one of the first attempts to evaluate the utility of zero-shot transfer for the task of node regression.

Blaž Škrlj, Sašo Džeroski, Nada Lavrač et al.

Feature ranking has been widely adopted in machine learning applications such as high-throughput biology and social sciences. The approaches of the popular Relief family of algorithms assign importances to features by iteratively accounting for nearest relevant and irrelevant instances. Despite their high utility, these algorithms can be computationally expensive and not-well suited for high-dimensional sparse input spaces. In contrast, recent embedding-based methods learn compact, low-dimensional representations, potentially facilitating down-stream learning capabilities of conventional learners. This paper explores how the Relief branch of algorithms can be adapted to benefit from (Riemannian) manifold-based embeddings of instance and target spaces, where a given embedding's dimensionality is intrinsic to the dimensionality of the considered data set. The developed ReliefE algorithm is faster and can result in better feature rankings, as shown by our evaluation on 20 real-life data sets for multi-class and multi-label classification tasks. The utility of ReliefE for high-dimensional data sets is ensured by its implementation that utilizes sparse matrix algebraic operations. Finally, the relation of ReliefE to other ranking algorithms is studied via the Fuzzy Jaccard Index.

Sebastian Mežnar, Nada Lavrač, Blaž Škrlj

Learning from complex real-life networks is a lively research area, with recent advances in learning information-rich, low-dimensional network node representations. However, state-of-the-art methods are not necessarily interpretable and are therefore not fully applicable to sensitive settings in biomedical or user profiling tasks, where explicit bias detection is highly relevant. The proposed SNoRe (Symbolic Node Representations) algorithm is capable of learning symbolic, human-understandable representations of individual network nodes, based on the similarity of neighborhood hashes which serve as features. SNoRe's interpretable features are suitable for direct explanation of individual predictions, which we demonstrate by coupling it with the widely used instance explanation tool SHAP to obtain nomograms representing the relevance of individual features for a given classification. To our knowledge, this is one of the first such attempts in a structural node embedding setting. In the experimental evaluation on eleven real-life datasets, SNoRe proved to be competitive to strong baselines, such as variational graph autoencoders, node2vec and LINE. The vectorized implementation of SNoRe scales to large networks, making it suitable for contemporary network learning and analysis tasks.

Matej Martinc, Blaž Škrlj, Sergej Pirkmajer et al.

The abundance of literature related to the widespread COVID-19 pandemic is beyond manual inspection of a single expert. Development of systems, capable of automatically processing tens of thousands of scientific publications with the aim to enrich existing empirical evidence with literature-based associations is challenging and relevant. We propose a system for contextualization of empirical expression data by approximating relations between entities, for which representations were learned from one of the largest COVID-19-related literature corpora. In order to exploit a larger scientific context by transfer learning, we propose a novel embedding generation technique that leverages SciBERT language model pretrained on a large multi-domain corpus of scientific publications and fine-tuned for domain adaptation on the CORD-19 dataset. The conducted manual evaluation by the medical expert and the quantitative evaluation based on therapy targets identified in the related work suggest that the proposed method can be successfully employed for COVID-19 therapy target discovery and that it outperforms the baseline FastText method by a large margin.

Nada Lavrač, Blaž Škrlj, Marko Robnik-Šikonja

Data preprocessing is an important component of machine learning pipelines, which requires ample time and resources. An integral part of preprocessing is data transformation into the format required by a given learning algorithm. This paper outlines some of the modern data processing techniques used in relational learning that enable data fusion from different input data types and formats into a single table data representation, focusing on the propositionalization and embedding data transformation approaches. While both approaches aim at transforming data into tabular data format, they use different terminology and task definitions, are perceived to address different goals, and are used in different contexts. This paper contributes a unifying framework that allows for improved understanding of these two data transformation techniques by presenting their unified definitions, and by explaining the similarities and differences between the two approaches as variants of a unified complex data transformation task. In addition to the unifying framework, the novelty of this paper is a unifying methodology combining propositionalization and embeddings, which benefits from the advantages of both in solving complex data transformation and learning tasks. We present two efficient implementations of the unifying methodology: an instance-based PropDRM approach, and a feature-based PropStar approach to data transformation and learning, together with their empirical evaluation on several relational problems. The results show that the new algorithms can outperform existing relational learners and can solve much larger problems.

Blaž Škrlj, Sašo Džeroski, Nada Lavrač et al.

Black-box neural network models are widely used in industry and science, yet are hard to understand and interpret. Recently, the attention mechanism was introduced, offering insights into the inner workings of neural language models. This paper explores the use of attention-based neural networks mechanism for estimating feature importance, as means for explaining the models learned from propositional (tabular) data. Feature importance estimates, assessed by the proposed Self-Attention Network (SAN) architecture, are compared with the established ReliefF, Mutual Information and Random Forest-based estimates, which are widely used in practice for model interpretation. For the first time we conduct scale-free comparisons of feature importance estimates across algorithms on ten real and synthetic data sets to study the similarities and differences of the resulting feature importance estimates, showing that SANs identify similar high-ranked features as the other methods. We demonstrate that SANs identify feature interactions which in some cases yield better predictive performance than the baselines, suggesting that attention extends beyond interactions of just a few key features and detects larger feature subsets relevant for the considered learning task.

Blaz Škrlj, Jan Kralj, Nada Lavrač

Relational data mining is becoming ubiquitous in many fields of study. It offers insights into behaviour of complex, real-world systems which cannot be modeled directly using propositional learning. We propose Symbolic Graph Embedding (SGE), an algorithm aimed to learn symbolic node representations. Built on the ideas from the field of inductive logic programming, SGE first samples a given node's neighborhood and interprets it as a transaction database, which is used for frequent pattern mining to identify logical conjuncts of items that co-occur frequently in a given context. Such patterns are in this work used as features to represent individual nodes, yielding interpretable, symbolic node embeddings. The proposed SGE approach on a venue classification task outperforms shallow node embedding methods such as DeepWalk, and performs similarly to metapath2vec, a black-box representation learner that can exploit node and edge types in a given graph. The proposed SGE approach performs especially well when small amounts of data are used for learning, scales to graphs with millions of nodes and edges, and can be run on an of-the-shelf laptop.

Blaž Škrlj, Jan Kralj, Nada Lavrač

Mining complex data in the form of networks is of increasing interest in many scientific disciplines. Network communities correspond to densely connected subnetworks, and often represent key functional parts of real-world systems. In this work, we propose Silhouette Community Detection (SCD), an approach for detecting communities, based on clustering of network node embeddings, i.e. real valued representations of nodes derived from their neighborhoods. We investigate the performance of the proposed SCD approach on 234 synthetic networks, as well as on a real-life social network. Even though SCD is not based on any form of modularity optimization, it performs comparably or better than state-of-the-art community detection algorithms, such as the InfoMap and Louvain algorithms. Further, we demonstrate how SCD's outputs can be used along with domain ontologies in semantic subgroup discovery, yielding human-understandable explanations of communities detected in a real-life protein interaction network. Being embedding-based, SCD is widely applicable and can be tested out-of-the-box as part of many existing network learning and exploration pipelines.

Blaž Škrlj, Jan Kralj, Janez Konc et al.

Network node embedding is an active research subfield of complex network analysis. This paper contributes a novel approach to learning network node embeddings and direct node classification using a node ranking scheme coupled with an autoencoder-based neural network architecture. The main advantages of the proposed Deep Node Ranking (DNR) algorithm are competitive or better classification performance, significantly higher learning speed and lower space requirements when compared to state-of-the-art approaches on 15 real-life node classification benchmarks. Furthermore, it enables exploration of the relationship between symbolic and the derived sub-symbolic node representations, offering insights into the learned node space structure. To avoid the space complexity bottleneck in a direct node classification setting, DNR computes stationary distributions of personalized random walks from given nodes in mini-batches, scaling seamlessly to larger networks. The scaling laws associated with DNR were also investigated on 1488 synthetic Erdős-Rényi networks, demonstrating its scalability to tens of millions of links.

Blaž Škrlj, Matej Martinc, Jan Kralj et al.

The use of background knowledge is largely unexploited in text classification tasks. This paper explores word taxonomies as means for constructing new semantic features, which may improve the performance and robustness of the learned classifiers. We propose tax2vec, a parallel algorithm for constructing taxonomy-based features, and demonstrate its use on six short text classification problems: prediction of gender, personality type, age, news topics, drug side effects and drug effectiveness. The constructed semantic features, in combination with fast linear classifiers, tested against strong baselines such as hierarchical attention neural networks, achieves comparable classification results on short text documents. The algorithm's performance is also tested in a few-shot learning setting, indicating that the inclusion of semantic features can improve the performance in data-scarce situations. The tax2vec capability to extract corpus-specific semantic keywords is also demonstrated. Finally, we investigate the semantic space of potential features, where we observe a similarity with the well known Zipf's law.

Matej Mihelčić, Goran Šimić, Mirjana Babić Leko et al.

We used redescription mining to find interpretable rules revealing associations between those determinants that provide insights about the Alzheimer's disease (AD). We extended the CLUS-RM redescription mining algorithm to a constraint-based redescription mining (CBRM) setting, which enables several modes of targeted exploration of specific, user-constrained associations. Redescription mining enabled finding specific constructs of clinical and biological attributes that describe many groups of subjects of different size, homogeneity and levels of cognitive impairment. We confirmed some previously known findings. However, in some instances, as with the attributes: testosterone, the imaging attribute Spatial Pattern of Abnormalities for Recognition of Early AD, as well as the levels of leptin and angiopoietin-2 in plasma, we corroborated previously debatable findings or provided additional information about these variables and their association with AD pathogenesis. Applying redescription mining on ADNI data resulted with the discovery of one largely unknown attribute: the Pregnancy-Associated Protein-A (PAPP-A), which we found highly associated with cognitive impairment in AD. Statistically significant correlations (p <= 0.01) were found between PAPP-A and various different clinical tests. The high importance of this finding lies in the fact that PAPP-A is a metalloproteinase, known to cleave insulin-like growth factor binding proteins. Since it also shares similar substrates with A Disintegrin and the Metalloproteinase family of enzymes that act as α-secretase to physiologically cleave amyloid precursor protein (APP) in the non-amyloidogenic pathway, it could be directly involved in the metabolism of APP very early during the disease course. Therefore, further studies should investigate the role of PAPP-A in the development of AD more thoroughly.

Matej Mihelčić, Sašo Džeroski, Nada Lavrač et al.

Redescription mining is a field of knowledge discovery that aims at finding different descriptions of similar subsets of instances in the data. These descriptions are represented as rules inferred from one or more disjoint sets of attributes, called views. As such, they support knowledge discovery process and help domain experts in formulating new hypotheses or constructing new knowledge bases and decision support systems. In contrast to previous approaches that typically create one smaller set of redescriptions satisfying a pre-defined set of constraints, we introduce a framework that creates large and heterogeneous redescription set from which user/expert can extract compact sets of differing properties, according to its own preferences. Construction of large and heterogeneous redescription set relies on CLUS-RM algorithm and a novel, conjunctive refinement procedure that facilitates generation of larger and more accurate redescription sets. The work also introduces the variability of redescription accuracy when missing values are present in the data, which significantly extends applicability of the method. Crucial part of the framework is the redescription set extraction based on heuristic multi-objective optimization procedure that allows user to define importance levels towards one or more redescription quality criteria. We provide both theoretical and empirical comparison of the novel framework against current state of the art redescription mining algorithms and show that it represents more efficient and versatile approach for mining redescriptions from data.