Ingo Scholtes

Vincenzo Perri, Lisi Qarkaxhija, Albin Zehe et al.

Natural Language Processing and Machine Learning have considerably advanced Computational Literary Studies. Similarly, the construction of co-occurrence networks of literary characters, and their analysis using methods from social network analysis and network science, have provided insights into the micro- and macro-level structure of literary texts. Combining these perspectives, in this work we study character networks extracted from a text corpus of J.R.R. Tolkien's Legendarium. We show that this perspective helps us to analyse and visualise the narrative style that characterises Tolkien's works. Addressing character classification, embedding and co-occurrence prediction, we further investigate the advantages of state-of-the-art Graph Neural Networks over a popular word embedding method. Our results highlight the large potential of graph learning in Computational Literary Studies.

Lisi Qarkaxhija, Vincenzo Perri, Ingo Scholtes

We introduce De Bruijn Graph Neural Networks (DBGNNs), a novel time-aware graph neural network architecture for time-resolved data on dynamic graphs. Our approach accounts for temporal-topological patterns that unfold in the causal topology of dynamic graphs, which is determined by causal walks, i.e. temporally ordered sequences of links by which nodes can influence each other over time. Our architecture builds on multiple layers of higher-order De Bruijn graphs, an iterative line graph construction where nodes in a De Bruijn graph of order k represent walks of length k-1, while edges represent walks of length k. We develop a graph neural network architecture that utilizes De Bruijn graphs to implement a message passing scheme that follows a non-Markovian dynamics, which enables us to learn patterns in the causal topology of a dynamic graph. Addressing the issue that De Bruijn graphs with different orders k can be used to model the same data set, we further apply statistical model selection to determine the optimal graph topology to be used for message passing. An evaluation in synthetic and empirical data sets suggests that DBGNNs can leverage temporal patterns in dynamic graphs, which substantially improves the performance in a supervised node classification task.

Christopher Blöcker, Chester Tan, Ingo Scholtes

Community detection is an essential tool for unsupervised data exploration and revealing the organisational structure of networked systems. With a long history in network science, community detection typically relies on objective functions, optimised with custom-tailored search algorithms, but often without leveraging recent advances in deep learning. Recently, first works have started incorporating such objectives into loss functions for deep graph clustering and pooling. We consider the map equation, a popular information-theoretic objective function for unsupervised community detection, and express it in differentiable tensor form for optimisation through gradient descent. Our formulation turns the map equation compatible with any neural network architecture, enables end-to-end learning, incorporates node features, and chooses the optimal number of clusters automatically, all without requiring explicit regularisation. Applied to unsupervised graph clustering tasks, we achieve competitive performance against state-of-the-art deep graph clustering baselines in synthetic and real-world datasets.

Franziska Heeg, Ingo Scholtes

Node centralities play a pivotal role in network science, social network analysis, and recommender systems. In temporal data, static path-based centralities like closeness or betweenness can give misleading results about the true importance of nodes in a temporal graph. To address this issue, temporal generalizations of betweenness and closeness have been defined that are based on the shortest time-respecting paths between pairs of nodes. However, a major issue of those generalizations is that the calculation of such paths is computationally expensive. Addressing this issue, we study the application of De Bruijn Graph Neural Networks (DBGNN), a time-aware graph neural network architecture, to predict temporal path-based centralities in time series data. We experimentally evaluate our approach in 13 temporal graphs from biological and social systems and show that it considerably improves the prediction of betweenness and closeness centrality compared to (i) a static Graph Convolutional Neural Network, (ii) an efficient sampling-based approximation technique for temporal betweenness, and (iii) two state-of-the-art time-aware graph learning techniques for dynamic graphs.

Christoph Gote, Ingo Scholtes, Frank Schweitzer

Data from software repositories have become an important foundation for the empirical study of software engineering processes. A recurring theme in the repository mining literature is the inference of developer networks capturing e.g. collaboration, coordination, or communication from the commit history of projects. Most of the studied networks are based on the co-authorship of software artefacts. Because this neglects detailed information on code changes and code ownership we introduce git2net, a scalable python software that facilitates the extraction of fine-grained co-editing networks in large git repositories. It uses text mining techniques to analyse the detailed history of textual modifications within files. We apply our tool in two case studies using GitHub repositories of multiple Open Source as well as a commercial software project. Specifically, we use data on more than 1.2 million commits and more than 25'000 developers to test a hypothesis on the relation between developer productivity and co-editing patterns in software teams. We argue that git2net opens up a massive new source of high-resolution data on human collaboration patterns that can be used to advance theory in empirical software engineering, computational social science, and organisational studies.

Christoph Gote, Ingo Scholtes, Frank Schweitzer

Data from software repositories have become an important foundation for the empirical study of software engineering processes. A recurring theme in the repository mining literature is the inference of developer networks capturing e.g. collaboration, coordination, or communication from the commit history of projects. Most of the studied networks are based on the co-authorship of software artefacts defined at the level of files, modules, or packages. While this approach has led to insights into the social aspects of software development, it neglects detailed information on code changes and code ownership, e.g. which exact lines of code have been authored by which developers, that is contained in the commit log of software projects. Addressing this issue, we introduce git2net, a scalable python software that facilitates the extraction of fine-grained co-editing networks in large git repositories. It uses text mining techniques to analyse the detailed history of textual modifications within files. This information allows us to construct directed, weighted, and time-stamped networks, where a link signifies that one developer has edited a block of source code originally written by another developer. Our tool is applied in case studies of an Open Source and a commercial software project. We argue that it opens up a massive new source of high-resolution data on human collaboration patterns.

Marcelo Serrano Zanetti, Ingo Scholtes, Claudio Juan Tessone et al.

Social organization and division of labor crucially influence the performance of collaborative software engineering efforts. In this paper, we provide a quantitative analysis of the relation between social organization and performance in Gentoo, an Open Source community developing a Linux distribution. We study the structure and dynamics of collaborations as recorded in the project's bug tracking system over a period of ten years. We identify a period of increasing centralization after which most interactions in the community were mediated by a single central contributor. In this period of maximum centralization, the central contributor unexpectedly left the project, thus posing a significant challenge for the community. We quantify how the rise, the activity as well as the subsequent sudden dropout of this central contributor affected both the social organization and the bug handling performance of the Gentoo community. We analyze social organization from the perspective of network theory and augment our quantitative findings by interviews with prominent members of the Gentoo community which shared their personal insights.

Marcelo Serrano Zanetti, Ingo Scholtes, Claudio Juan Tessone et al.

Efficient bug triaging procedures are an important precondition for successful collaborative software engineering projects. Triaging bugs can become a laborious task particularly in open source software (OSS) projects with a large base of comparably inexperienced part-time contributors. In this paper, we propose an efficient and practical method to identify valid bug reports which a) refer to an actual software bug, b) are not duplicates and c) contain enough information to be processed right away. Our classification is based on nine measures to quantify the social embeddedness of bug reporters in the collaboration network. We demonstrate its applicability in a case study, using a comprehensive data set of more than 700,000 bug reports obtained from the Bugzilla installation of four major OSS communities, for a period of more than ten years. For those projects that exhibit the lowest fraction of valid bug reports, we find that the bug reporters' position in the collaboration network is a strong indicator for the quality of bug reports. Based on this finding, we develop an automated classification scheme that can easily be integrated into bug tracking platforms and analyze its performance in the considered OSS communities. A support vector machine (SVM) to identify valid bug reports based on the nine measures yields a precision of up to 90.3% with an associated recall of 38.9%. With this, we significantly improve the results obtained in previous case studies for an automated early identification of bugs that are eventually fixed. Furthermore, our study highlights the potential of using quantitative measures of social organization in collaborative software engineering. It also opens a broad perspective for the integration of social awareness in the design of support infrastructures.

Marcelo Serrano Zanetti, Emre Sarigol, Ingo Scholtes et al.

The success of open source projects crucially depends on the voluntary contributions of a sufficiently large community of users. Apart from the mere size of the community, interesting questions arise when looking at the evolution of structural features of collaborations between community members. In this article, we discuss several network analytic proxies that can be used to quantify different aspects of the social organisation in social collaboration networks. We particularly focus on measures that can be related to the cohesiveness of the communities, the distribution of responsibilities and the resilience against turnover of community members. We present a comparative analysis on a large-scale dataset that covers the full history of collaborations between users of 14 major open source software communities. Our analysis covers both aggregate and time-evolving measures and highlights differences in the social organisation across communities. We argue that our results are a promising step towards the definition of suitable, potentially multi-dimensional, resilience and risk indicators for open source software communities.

Jan von Pichowski, Alžbeta Hrabošová, Ingo Scholtes et al.

Graph pooling is commonly applied in graph classification, yet its empirical gains over standard WL-1 expressive GNNs are often marginal or inconsistent. We study this gap by analysing the interaction between node features and graph topology and their effect on pooling objectives. Our analysis reveals that pooling operators require node features that are well-aligned with the graph's topology -- a condition often overlooked and not guaranteed in empirical networks. We formalise fundamental requirements for node features to enable effective pooling, and introduce a quantitative measure of feature quality. Our empirical evaluation shows that, when these requirements are satisfied, pooling can be beneficial and improve performance on appropriate datasets.

Jan von Pichowski, Christopher Blöcker, Ingo Scholtes

Graph pooling compresses graphs and summarises their topological properties and features in a vectorial representation. It is an essential part of deep graph representation learning and is indispensable in graph-level tasks like classification or regression. Current approaches pool hierarchical structures in graphs by iteratively applying shallow pooling operators up to a fixed depth. However, they disregard the interdependencies between structures at different hierarchical levels and do not adapt to datasets that contain graphs with different sizes that may require pooling with various depths. To address these issues, we propose MDL-Pool, a pooling operator based on the minimum description length (MDL) principle, whose loss formulation explicitly models the interdependencies between different hierarchical levels and facilitates a direct comparison between multiple pooling alternatives with different depths. MDP-Pool builds on the map equation, an information-theoretic objective function for community detection, which naturally implements Occam's razor and balances between model complexity and goodness-of-fit via the MDL. We demonstrate MDL-Pool's competitive performance in an empirical evaluation against various baselines across standard graph classification datasets.

Matthias Nägele, Cedric Bös, Chester Tan et al.

General-relativistic magnetohydrodynamic (GR-MHD) simulations are essential for studying black hole accretion, relativistic jets, and magnetic reconnection, yet their computational cost severely limits systematic parameter exploration. We investigate neural operator surrogates for two astrophysically relevant simulation scenarios produced by the Black Hole Accretion Code (\texttt{BHAC}). First, a Physics Informed Fourier Neural Operator (PINO) is trained on the special-relativistic resistive MHD (SRRMHD) evolution of the Orszag-Tang vortex over a range of resistivities spanning the Sweet-Parker and fast reconnection regimes. By embedding the governing equations as an additional loss term evaluated at finer temporal resolution than the available data supervision, the model learns dynamics at time steps where no simulation data is provided, enabling recovery of plasmoid formation that a data-only baseline trained on the same sparse snapshots fails to reproduce. To our knowledge, the present work is the first application of a physics informed neural operator to special relativistic resistive MHD, and the first to investigate the capability of such models to resolve plasmoid formation in SRRMHD. In a second line of investigation, an OFormer-style Transformer Neural Operator is trained on the evolution of spine-sheath relativistic jets created with \texttt{BHAC}, in special-relativistic MHD (SRMHD). The model is directly applied on the adaptive mesh, highlighting the need for linear attention due to long sequences. The neural surrogate model is capable of capturing most of the major details, especially in early predictions. To our knowledge, this constitutes the first application of a neural operator directly on a high resolution adaptive mesh refinement grid in the context of MHD simulations.

Moritz Lampert, Ingo Scholtes

Many Graph Neural Networks (GNNs) add self-loops to a graph to include feature information about a node itself at each layer. However, if the GNN consists of more than one layer, this information can return to its origin via cycles in the graph topology. Intuition suggests that this "backflow" of information should be larger in graphs with self-loops compared to graphs without. In this work, we counter this intuition and show that for certain GNN architectures, the information a node gains from itself can be smaller in graphs with self-loops compared to the same graphs without. We adopt an analytical approach for the study of statistical graph ensembles with a given degree sequence and show that this phenomenon, which we call the self-loop paradox, can depend both on the number of GNN layers $k$ and whether $k$ is even or odd. We experimentally validate our theoretical findings in a synthetic node classification task and investigate its practical relevance in 23 real-world graphs.

Christopher Blöcker, Martin Rosvall, Ingo Scholtes et al.

Deep graph learning and network science both analyze graphs but approach similar problems from different perspectives. Whereas network science focuses on models and measures that reveal the organizational principles of complex systems with explicit assumptions, deep graph learning focuses on flexible and generalizable models that learn patterns in graph data in an automated fashion. Despite these differences, both fields share the same goal: to better model and understand patterns in graph-structured data. Early efforts to integrate methods, models, and measures from network science and deep graph learning indicate significant untapped potential. In this position, we explore opportunities at their intersection. We discuss open challenges in deep graph learning, including data augmentation, improved evaluation practices, higher-order models, and pooling methods. Likewise, we highlight challenges in network science, including scaling to massive graphs, integrating continuous gradient-based optimization, and developing standardized benchmarks.

Franziska Heeg, Jonas Sauer, Petra Mutzel et al.

An important characteristic of temporal graphs is how the directed arrow of time influences their causal topology, i.e., which nodes can possibly influence each other causally via time-respecting paths. The resulting patterns are often neglected by temporal graph neural networks (TGNNs). To formally analyze the expressive power of TGNNs, we lack a generalization of graph isomorphism to temporal graphs that fully captures their causal topology. Addressing this gap, we introduce the notion of consistent event graph isomorphism, which utilizes a time-unfolded representation of time-respecting paths in temporal graphs. We compare this definition with existing notions of temporal graph isomorphisms. We illustrate and highlight the advantages of our approach and develop a temporal generalization of the Weisfeiler-Leman algorithm to heuristically distinguish non-isomorphic temporal graphs. Building on this theoretical foundation, we derive a novel message passing scheme for temporal graph neural networks that operates on the event graph representation of temporal graphs. An experimental evaluation shows that our approach performs well in a temporal graph classification experiment.

Lisi Qarkaxhija, Anatol E. Wegner, Ingo Scholtes

Message passing neural networks (MPNNs) operate on graphs by exchanging information between neigbouring nodes. MPNNs have been successfully applied to various node-, edge-, and graph-level tasks in areas like molecular science, computer vision, natural language processing, and combinatorial optimization. However, most MPNNs require training on large amounts of labeled data, which can be costly and time-consuming. In this work, we explore the use of various untrained message passing layers in graph neural networks, i.e. variants of popular message passing architecture where we remove all trainable parameters that are used to transform node features in the message passing step. Focusing on link prediction, we find that untrained message passing layers can lead to competitive and even superior performance compared to fully trained MPNNs, especially in the presence of high-dimensional features. We provide a theoretical analysis of untrained message passing by relating the inner products of features implicitly produced by untrained message passing layers to path-based topological node similarity measures. As such, untrained message passing architectures can be viewed as a highly efficient and interpretable approach to link prediction.

Jan von Pichowski, Vincenzo Perri, Lisi Qarkaxhija et al.

The modelling of temporal patterns in dynamic graphs is an important current research issue in the development of time-aware GNNs. Whether or not a specific sequence of events in a temporal graph constitutes a temporal pattern not only depends on the frequency of its occurrence. We consider whether it deviates from what is expected in a temporal graph where timestamps are randomly shuffled. While accounting for such a random baseline is important to model temporal patterns, it has mostly been ignored by current temporal graph neural networks. To address this issue we propose HYPA-DBGNN, a novel two-step approach that combines (i) the inference of anomalous sequential patterns in time series data on graphs based on a statistically principled null model, with (ii) a neural message passing approach that utilizes a higher-order De Bruijn graph whose edges capture overrepresented sequential patterns. Our method leverages hypergeometric graph ensembles to identify anomalous edges within both first- and higher-order De Bruijn graphs, which encode the temporal ordering of events. The model introduces an inductive bias that enhances model interpretability. We evaluate our approach for static node classification using benchmark datasets and a synthetic dataset that showcases its ability to incorporate the observed inductive bias regarding over- and under-represented temporal edges. We demonstrate the framework's effectiveness in detecting similar patterns within empirical datasets, resulting in superior performance compared to baseline methods in node classification tasks. To the best of our knowledge, our work is the first to introduce statistically informed GNNs that leverage temporal and causal sequence anomalies. HYPA-DBGNN represents a path for bridging the gap between statistical graph inference and neural graph representation learning, with potential applications to static GNNs.

Moritz Lampert, Christopher Blöcker, Ingo Scholtes

Dynamic link prediction is an important problem considered in many recent works that propose approaches for learning temporal edge patterns. To assess their efficacy, models are evaluated on continuous-time and discrete-time temporal graph datasets, typically using a traditional batch-oriented evaluation setup. However, as we show in this work, a batch-oriented evaluation is often unsuitable and can cause several issues. Grouping edges into fixed-sized batches regardless of their occurrence time leads to information loss or leakage, depending on the temporal granularity of the data. Furthermore, fixed-size batches create time windows with different durations, resulting in an inconsistent dynamic link prediction task. In this work, we empirically show how traditional batch-based evaluation leads to skewed model performance and hinders the fair comparison of methods. We mitigate this problem by reformulating dynamic link prediction as a link forecasting task that better accounts for temporal information present in the data.

Christoph Gote, Pavlin Mavrodiev, Frank Schweitzer et al.

Massive data from software repositories and collaboration tools are widely used to study social aspects in software development. One question that several recent works have addressed is how a software project's size and structure influence team productivity, a question famously considered in Brooks' law. Recent studies using massive repository data suggest that developers in larger teams tend to be less productive than smaller teams. Despite using similar methods and data, other studies argue for a positive linear or even super-linear relationship between team size and productivity, thus contesting the view of software economics that software projects are diseconomies of scale. In our work, we study challenges that can explain the disagreement between recent studies of developer productivity in massive repository data. We further provide, to the best of our knowledge, the largest, curated corpus of GitHub projects tailored to investigate the influence of team size and collaboration patterns on individual and collective productivity. Our work contributes to the ongoing discussion on the choice of productivity metrics in the operationalisation of hypotheses about determinants of successful software projects. It further highlights general pitfalls in big data analysis and shows that the use of bigger data sets does not automatically lead to more reliable insights.

Unai Alvarez-Rodriguez, Luka V. Petrović, Ingo Scholtes

We introduce time-ordered multibody interactions to describe complex systems manifesting temporal as well as multibody dependencies. First, we show how the dynamics of multivariate Markov chains can be decomposed in ensembles of time-ordered multibody interactions. Then, we present an algorithm to extract those interactions from data capturing the system-level dynamics of node states and a measure to characterize the complexity of interaction ensembles. Finally, we experimentally validate the robustness of our algorithm against statistical errors and its efficiency at inferring parsimonious interaction ensembles.

Christoph Gote, Vincenzo Perri, Ingo Scholtes

Networks are frequently used to model complex systems comprised of interacting elements. While edges capture the topology of direct interactions, the true complexity of many systems originates from higher-order patterns in paths by which nodes can indirectly influence each other. Path data, representing ordered sequences of consecutive direct interactions, can be used to model these patterns. On the one hand, to avoid overfitting, such models should only consider those higher-order patterns for which the data provide sufficient statistical evidence. On the other hand, we hypothesise that network models, which capture only direct interactions, underfit higher-order patterns present in data. Consequently, both approaches are likely to misidentify influential nodes in complex networks. We contribute to this issue by proposing five centrality measures based on MOGen, a multi-order generative model that accounts for all indirect influences up to a maximum distance but disregards influences at higher distances. We compare MOGen-based centralities to equivalent measures for network models and path data in a prediction experiment where we aim to identify influential nodes in out-of-sample data. Our results show strong evidence supporting our hypothesis. MOGen consistently outperforms both the network model and path-based prediction. We further show that the performance difference between MOGen and the path-based approach disappears if we have sufficient observations, confirming that the error is due to overfitting.

Christoph Gote, Giona Casiraghi, Frank Schweitzer et al.

We propose a novel sequence prediction method for sequential data capturing node traversals in graphs. Our method builds on a statistical modelling framework that combines multiple higher-order network models into a single multi-order model. We develop a technique to fit such multi-order models in empirical sequential data and to select the optimal maximum order. Our framework facilitates both next-element and full sequence prediction given a sequence-prefix of any length. We evaluate our model based on six empirical data sets containing sequences from website navigation as well as public transport systems. The results show that our method out-performs state-of-the-art algorithms for next-element prediction. We further demonstrate the accuracy of our method during out-of-sample sequence prediction and validate that our method can scale to data sets with millions of sequences.

Luka V. Petrović, Ingo Scholtes

We study the problem of learning the Markov order in categorical sequences that represent paths in a network, i.e. sequences of variable lengths where transitions between states are constrained to a known graph. Such data pose challenges for standard Markov order detection methods and demand modelling techniques that explicitly account for the graph constraint. Adopting a multi-order modelling framework for paths, we develop a Bayesian learning technique that (i) more reliably detects the correct Markov order compared to a competing method based on the likelihood ratio test, (ii) requires considerably less data compared to methods using AIC or BIC, and (iii) is robust against partial knowledge of the underlying constraints. We further show that a recently published method that uses a likelihood ratio test has a tendency to overfit the true Markov order of paths, which is not the case for our Bayesian technique. Our method is important for data scientists analyzing patterns in categorical sequence data that are subject to (partially) known constraints, e.g. sequences with forbidden words, mobility trajectories and click stream data, or sequence data in bioinformatics. Addressing the key challenge of model selection, our work is further relevant for the growing body of research that emphasizes the need for higher-order models in network analysis.

Vincenzo Perri, Ingo Scholtes

Network visualisation techniques are important tools for the exploratory analysis of complex systems. While these methods are regularly applied to visualise data on complex networks, we increasingly have access to time series data that can be modelled as temporal networks or dynamic graphs. In dynamic graphs, the temporal ordering of time-stamped edges determines the causal topology of a system, i.e., which nodes can, directly and indirectly, influence each other via a so-called causal path. This causal topology is crucial to understand dynamical processes, assess the role of nodes, or detect clusters. However, we lack graph drawing techniques that incorporate this information into static visualisations. Addressing this gap, we present a novel dynamic graph visualisation algorithm that utilises higher-order graphical models of causal paths in time series data to compute time-aware static graph visualisations. These visualisations combine the simplicity and interpretability of static graphs with a time-aware layout algorithm that highlights patterns in the causal topology that result from the temporal dynamics of edges.