Abstract High‐dimensional data analysis often uses dimensionality reduction (DR, also called projection) to map data patterns to human‐digestible visual patterns in a 2D scatterplot. Yet, DR methods may fail to show true data patterns and/or create visual patterns that do not represent any data patterns. Projection Quality Metrics (PQMs) are used as objective measures to gauge the above process: the higher a projection's scores in PQMs, the more it is deemed faithful to the data it represents. We show that, while PQMs can be used as exclusion criteria — low values usually mean poor projections — the converse does not always hold. For this, we develop a technique to automatically generate projections that score similar or even higher PQM values than projections created by well‐known techniques, but show different, often confusing, visual patterns. Our results show that accepted PQMs cannot be used as an exclusive way to tell whether a projection yields accurate and interpretable visual patterns — in this sense, PQMs play a role akin to that of summary statistics in exploratory data analysis. We also show that not all studied metrics can befooled equally well, suggesting a ranking of metrics in their ability to reliably capture quality.

The semantic similarity between documents of a text corpus can be visualized using map-like metaphors based on two-dimensional scatterplot layouts. These layouts result from a dimensionality reduction on the document-term matrix or a representation within a latent embedding, including topic models. Thereby, the resulting layout depends on the input data and hyperparameters of the dimensionality reduction and is therefore affected by changes in them. Furthermore, the resulting layout is affected by changes in the input data and hyperparameters of the dimensionality reduction. However, such changes to the layout require additional cognitive efforts from the user. In this work, we present a sensitivity study that analyzes the stability of these layouts concerning (1) changes in the text corpora, (2) changes in the hyperparameter, and (3) randomness in the initialization. Our approach has two stages: data measurement and data analysis. First, we derived layouts for the combination of three text corpora and six text embeddings and a grid-search-inspired hyperparameter selection of the dimensionality reductions. Afterward, we quantified the similarity of the layouts through ten metrics, concerning local and global structures and class separation. Second, we analyzed the resulting 42 817 tabular data points in a descriptive statistical analysis. From this, we derived guidelines for informed decisions on the layout algorithm and highlight specific hyperparameter settings. We provide our implementation as a Git repository at hpicgs/Topic-Models-and-Dimensionality-Reduction-Sensitivity-Study and results as Zenodo archive at DOI:10.5281/zenodo.12772898.

Multidimensional projections are effective techniques for depicting high-dimensional data. The point patterns created by such techniques, or a technique’s visual signature , depend — apart from the data themselves — on the technique design and its parameter settings. Controlling such visual signatures — something that only few projections allow — can bring additional freedom for generating insightful depictions of the data. We present a novel projection technique — ShaRP — that allows explicit control on such visual signatures in terms of shapes of similar-value point clusters (settable to rectangles, triangles, ellipses, and convex polygons) and the projection space (2D or 3D Euclidean or S 2 ). We show that ShaRP scales computationally well with dimensionality and dataset size, provides its signature-control by a small set of parameters, allows trading off projection quality to signature enforcement, and can be used to generate decision maps to explore the behavior of trained machine-learning classifiers.

Abstract Explaining Machine Learning (ML) — and especially Deep Learning (DL) — classifiers' decisions is a subject of interest across fields due to the increasing ubiquity of such models in computing systems. As models get increasingly complex, relying on sophisticated machinery to recognize data patterns, explaining their behavior becomes more difficult. Directly visualizing classifier behavior is in general infeasible, as they create partitions of the data space, which is typically high dimensional. In recent years, Decision Boundary Maps (DBMs) have been developed, taking advantage of projection and inverse projection techniques. By being able to map 2D points back to the data space and subsequently run a classifier, DBMs represent a slice of classifier outputs. However, we recognize that DBMs without additional explanatory views are limited in their applicability. In this work, we propose augmenting the naive DBM generating process with views that provide more in‐depth information about classifier behavior, such as whether the training procedure is locally stable. We describe our proposed views — which we term Differentiable Decision Boundary Maps — over a running example, explaining how our work enables drawing new and useful conclusions from these dense maps. We further demonstrate the value of these conclusions by showing how useful they would be in carrying out or preventing a dataset poisoning attack. We thus provide evidence of the ability of our proposed views to make DBMs significantly more trustworthy and interpretable, increasing their utility as a model understanding tool.

Relational information between different types of entities is often modelled by a multilayer network (MLN) - a network with subnetworks represented by layers. The layers of an MLN can be arranged in different ways in a visual representation, however, the impact of the arrangement on the readability of the network is an open question. Therefore, we studied this impact for several commonly occurring tasks related to MLN analysis. Additionally, layer arrangements with a dimensionality beyond 2D, which are common in this scenario, motivate the use of stereoscopic displays. We ran a human subject study utilising a Virtual Reality headset to evaluate 2D, 2.5D, and 3D layer arrangements. The study employs six analysis tasks that cover the spectrum of an MLN task taxonomy, from path finding and pattern identification to comparisons between and across layers. We found no clear overall winner. However, we explore the task-to-arrangement space and derive empirical-based recommendations on the effective use of 2D, 2.5D, and 3D layer arrangements for MLNs.

We present SAXRegEx, a method for pattern search in multivariate time series in the presence of various distortions, such as duration variation, warping, and time delay between signals. For example, in the automotive industry, calibration engineers spontaneously search for event-induced patterns in new measurements under time pressure. Current methods do not sufficiently address duration (horizontal along the time axis) scaling and inter-track time delay. One reason is that it can be overwhelmingly complex to consider scaling and warping jointly and analyze temporal dynamics and attribute interrelation simultaneously. SAXRegEx meets this challenge with a novel symbolic representation adapted to handle time series with multiple tracks. We employ methods from text retrieval, i.e., regular expression matching, to perform a pattern retrieval and develop a novel query expansion algorithm to deal flexibly with pattern distortions. Experiments show the effectiveness of our approach, especially in the presence of such distortions, and its efficiency surpassing the benchmarking methods. We have developed a user interface with the emphasis on multivariate query definition with inter-track time shifts and algorithm parameter setting. While we design the method primarily for automotive data, it is well transferable to other domains.

We present PSEUDo, a visual pattern retrieval tool for multivariate time series. It aims to overcome the uneconomic (re-)training problem accompanying deep learning-based methods. Very high-dimensional time series emerge on an unprecedented scale due to increasing sensor usage and data storage. Visual pattern search is one of the most frequent tasks on time series. Automatic pattern retrieval methods often suffer from inefficient training data, a lack of ground truth labels, and a discrepancy between the similarity perceived by the algorithm and required by the user or the task. Our proposal is based on the query-aware locality-sensitive hashing technique to create a representation of multivariate time series windows. It features sub-linear training and inference time with respect to data dimensions. This performance gain allows an instantaneous relevance-feedback-driven adaption to converge to users' similarity notion. We demonstrate PSEUDo's performance in terms of accuracy, speed, steerability, and usability through quantitative benchmarks with representative time series retrieval methods and a case study. We find that PSEUDo detects patterns in high-dimensional time series efficiently, improves the result with relevance feedback through feature selection, and allows an understandable as well as user-friendly retrieval process.

Matrix representations are one of the main established and empirically proven to be effective visualization techniques for relational (or network) data. However, matrices-similar to node-link diagrams-are most effective if their layout reveals the underlying data topology. Given the many developed algorithms, a practical problem arises: "Which matrix reordering algorithm should I choose for my dataset at hand?" To make matters worse, different reordering algorithms applied to the same dataset may let significantly different visual matrix patterns emerge. This leads to the question of trustworthiness and explainability of these fully automated, often heuristic, black-box processes. We present GUIRO, a Visual Analytics system that helps novices, network analysts, and algorithm designers to open the black-box. Users can investigate the usefulness and expressiveness of 70 accessible matrix reordering algorithms. For network analysts, we introduce a novel model space representation and two interaction techniques for a user-guided reordering of rows or columns, and especially groups thereof (submatrix reordering). These novel techniques contribute to the understanding of the global and local dataset topology. We support algorithm designers by giving them access to 16 reordering quality metrics and visual exploration means for comparing reordering implementations on a row/column permutation level. We evaluated GUIRO in a guided explorative user study with 12 subjects, a case study demonstrating its usefulness in a real-world scenario, and through an expert study gathering feedback on our design decisions. We found that our proposed methods help even inexperienced users to understand matrix patterns and allow a user-guided steering of reordering algorithms. GUIRO helps to increase the transparency of matrix reordering algorithms, thus helping a broad range of users to get a better insight into the complex reordering process, in turn supporting data and reordering algorithm insights.

We present Scalable Insets, a technique for interactively exploring and navigating large numbers of annotated patterns in multiscale visualizations such as gigapixel images, matrices, or maps. Exploration of many but sparsely-distributed patterns in multiscale visualizations is challenging as visual representations change across zoom levels, context and navigational cues get lost upon zooming, and navigation is time consuming. Our technique visualizes annotated patterns too small to be identifiable at certain zoom levels using insets, i.e., magnified thumbnail views of the annotated patterns. Insets support users in searching, comparing, and contextualizing patterns while reducing the amount of navigation needed. They are dynamically placed either within the viewport or along the boundary of the viewport to offer a compromise between locality and context preservation. Annotated patterns are interactively clustered by location and type. They are visually represented as an aggregated inset to provide scalable exploration within a single viewport. In a controlled user study with 18 participants, we found that Scalable Insets can speed up visual search and improve the accuracy of pattern comparison at the cost of slower frequency estimation compared to a baseline technique. A second study with 6 experts in the field of genomics showed that Scalable Insets is easy to learn and provides first insights into how Scalable Insets can be applied in an open-ended data exploration scenario.

Five years after the first state-of-the-art report on Commercial Visual Analytics Systems we present a reevaluation of the Big Data Analytics field. We build on the success of the 2012 survey, which was influential even beyond the boundaries of the InfoVis and Visual Analytics (VA) community. While the field has matured significantly since the original survey, we find that innovation and research-driven development are increasingly sacrificed to satisfy a wide range of user groups. We evaluate new product versions on established evaluation criteria, such as available features, performance, and usability, to extend on and assure comparability with the previous survey. We also investigate previously unavailable products to paint a more complete picture of the commercial VA landscape. Furthermore, we introduce novel measures, like suitability for specific user groups and the ability to handle complex data types, and undertake a new case study to highlight innovative features. We explore the achievements in the commercial sector in addressing VA challenges and propose novel developments that should be on systems' roadmaps in the coming years.

Neural sequence-to-sequence models have proven to be accurate and robust for many sequence prediction tasks, and have become the standard approach for automatic translation of text. The models work with a five-stage blackbox pipeline that begins with encoding a source sequence to a vector space and then decoding out to a new target sequence. This process is now standard, but like many deep learning methods remains quite difficult to understand or debug. In this work, we present a visual analysis tool that allows interaction and "what if"-style exploration of trained sequence-to-sequence models through each stage of the translation process. The aim is to identify which patterns have been learned, to detect model errors, and to probe the model with counterfactual scenario. We demonstrate the utility of our tool through several real-world sequence-to-sequence use cases on large-scale models.

Abstract The visualization community has developed to date many intuitions and understandings of how to judge thequalityof views in visualizing data. The computation of a visualization's quality and usefulness ranges from measuring clutter and overlap, up to the existence and perception of specific (visual) patterns. This survey attempts to report, categorize and unify the diverse understandings and aims to establish a common vocabulary that will enable a wide audience to understand their differences and subtleties. For this purpose, we present a commonly applicable quality metric formalization that should detail and relate all constituting parts of a quality metric. We organize our corpus of reviewed research papers along the data types established in the information visualization community: multi‐ and high‐dimensional, relational, sequential, geospatial and text data. For each data type, we select the visualization subdomains in which quality metrics are an active research field and report their findings, reason on the underlying concepts, describe goals and outline the constraints and requirements. One central goal of this survey is to provide guidance on future research opportunities for the field and outline how different visualization communities could benefit from each other by applying or transferring knowledge to their respective subdomain. Additionally, we aim to motivate the visualization community to compare computed measures to the perception of humans.

Urban planning and intelligent transportation management are facing key challenges in today's ever more urbanized world. Providing the right tools to city planners is crucial to cope with these challenges. Data collected from citizens' mobile communication can be used as the foundation for such tools. These kinds of data can facilitate various analysis tasks, such as the extraction of human movement patterns or determining the urban dynamics of a city. City planners can closely monitor such patterns based on which strategic decisions can be taken to improve a city's infrastructure. In this paper, we introduce a novel visual analytics approach for pattern exploration and search in global system for mobile communications mobile networks. We define geospatial and matrix representations of data, which can be interactively navigated. The approach integrates data visualization with suitable data analysis algorithms, allowing to spatially and temporally compare mobile usage, identify regularities, as well as anomalies in daily mobility patterns across regions and user groups. As an extension to our visual analytics approach, we further introduce space-time prisms with uncertain markers to visually analyze the uncertainty of urban mobility patterns.

Clustering is a core building block for data analysis, aiming to extract otherwise hidden structures and relations from raw datasets, such as particular groups that can be effectively related, compared, and interpreted. A plethora of visual-interactive cluster analysis techniques has been proposed to date, however, arriving at useful clusterings often requires several rounds of user interactions to fine-tune the data preprocessing and algorithms. We present a multi-stage Visual Analytics (VA) approach for iterative cluster refinement together with an implementation (SOMFlow) that uses Self-Organizing Maps (SOM) to analyze time series data. It supports exploration by offering the analyst a visual platform to analyze intermediate results, adapt the underlying computations, iteratively partition the data, and to reflect previous analytical activities. The history of previous decisions is explicitly visualized within a flow graph, allowing to compare earlier cluster refinements and to explore relations. We further leverage quality and interestingness measures to guide the analyst in the discovery of useful patterns, relations, and data partitions. We conducted two pair analytics experiments together with a subject matter expert in speech intonation research to demonstrate that the approach is effective for interactive data analysis, supporting enhanced understanding of clustering results as well as the interactive process itself.

Many visualizations have proven to be effective in supporting various software related tasks. Although multiple media can be used to display a visualization, the standard computer screen is used the most. We hypothesize that the medium has a role in their effectiveness. We investigate our hypotheses by conducting a controlled user experiment. In the experiment we focus on the 3D city visualization technique used for software comprehension tasks. We deploy 3D city visualizations across a standard computer screen (SCS), an immersive 3D environment (I3D), and a physical 3D printed model (P3D). We asked twenty-seven participants (whom we divided in three groups for each medium) to visualize software systems of various sizes, solve a set of uniform comprehension tasks, and complete a questionnaire. We measured the effectiveness of visualizations in terms of performance, recollection, and user experience. We found that even though developers using P3D required the least time to identify outliers, they perceived the least difficulty when visualizing systems based on SCS. Moreover, developers using I3D obtained the highest recollection.

Abstract Named entity recognition (NER) denotes the task to detect entities and their corresponding classes, such as person or location, in unstructured text data. For most applications, state of the art NER software is producing reasonable results. However, as a consequence of the methodological limitations and the well‐known pitfalls when analyzing natural language data, the NER results are likely to contain ambiguities. In this paper, we present an interactive NER ambiguity resolution technique, which enables users to create (post‐processing) rules for named entity recognition data based on the content and entity context of the analyzed documents. We specifically address the problem that in use‐cases where ambiguities are problematic, such as the attribution of fictional characters with traits, it is often unfeasible to train models on custom data to improve state of the art NER software. We derive an iterative process model for improving NER results, show an interactive NER ambiguity resolution prototype, illustrate our approach with contemporary literature, and discuss our work and future research.

In this work we address the problem of retrieving potentially interesting matrix views to support the exploration of networks. We introduce Matrix Diagnostics (or Magnostics), following in spirit related approaches for rating and ranking other visualization techniques, such as Scagnostics for scatter plots. Our approach ranks matrix views according to the appearance of specific visual patterns, such as blocks and lines, indicating the existence of topological motifs in the data, such as clusters, bi-graphs, or central nodes. Magnostics can be used to analyze, query, or search for visually similar matrices in large collections, or to assess the quality of matrix reordering algorithms. While many feature descriptors for image analyzes exist, there is no evidence how they perform for detecting patterns in matrices. In order to make an informed choice of feature descriptors for matrix diagnostics, we evaluate 30 feature descriptors-27 existing ones and three new descriptors that we designed specifically for MAGNOSTICS-with respect to four criteria: pattern response, pattern variability, pattern sensibility, and pattern discrimination. We conclude with an informed set of six descriptors as most appropriate for Magnostics and demonstrate their application in two scenarios; exploring a large collection of matrices and analyzing temporal networks.

Abstract This survey provides a description of algorithms to reorder visual matrices of tabular data and adjacency matrix of Networks. The goal of this survey is to provide a comprehensive list of reordering algorithms published in different fields such as statistics, bioinformatics, or graph theory. While several of these algorithms are described in publications and others are available in software libraries and programs, there is little awareness of what is done across all fields. Our survey aims at describing these reordering algorithms in a unified manner to enable a wide audience to understand their differences and subtleties. We organize this corpus in a consistent manner, independently of the application or research field. We also provide practical guidance on how to select appropriate algorithms depending on the structure and size of the matrix to reorder, and point to implementations when available.

Abstract Matrix visualization is an established technique in the analysis of relational data. It is applicable to large, dense networks, where node‐link representations may not be effective. Recently, domains have emerged in which the comparative analysis of sets of matrices of potentially varying size is relevant. For example, to monitor computer network traffic a dynamic set of hosts and their peer‐to‐peer connections on different ports must be analysed. A matrix visualization focused on the display of one matrix at a time cannot cope with this task. We address the research problem of the visual analysis of sets of matrices. We present a technique for comparing matrices of potentially varying size. Our approach considers the rows and/or columns of a matrix as the basic elements of the analysis. We project these vectors for pairs of matrices into a low‐dimensional space which is used as the reference to compare matrices and identify relationships among them. Bipartite graph matching is applied on the projected elements to compute a measure of distance. A key advantage of this measure is that it can be interpreted and manipulated as a visual distance function, and serves as a comprehensible basis for ranking, clustering and comparison in sets of matrices. We present an interactive system in which users may explore the matrix distances and understand potential differences in a set of matrices. A flexible semantic zoom mechanism enables users to navigate through sets of matrices and identify patterns at different levels of detail. We demonstrate the effectiveness of our approach through a case study and provide a technical evaluation to illustrate its strengths.