Rankings are a popular and universal approach to structuring otherwise unorganized collections of items by computing rank for each item based on the value one or more its attributes. This allows us, example, prioritize tasks evaluate performance products relative other. While visualization ranking itself is straightforward, interpretation not, because an represents only summary potentially complicated relationship between attributes those other items. It also common that alternative rankings...

Visual analytics (VA) is typically applied in scenarios where complex data has to be analyzed. Unfortunately, there a natural correlation between the complexity of and tools study them. An adverse effect complicated that analytical goals are more difficult reach. Therefore, it makes sense consider methods guide or assist users visual analysis process. Several such already exist literature, yet we lacking general model facilitates in-depth reasoning about guidance. We establish by extending...

Understanding the recommendations of an artificial intelligence (AI) based assistant for decision-making is especially important in high-risk tasks, such as deciding whether a mushroom edible or poisonous. To foster user understanding and appropriate trust systems, we assessed effects explainable (XAI) methods educational intervention on AI-assisted behavior 2 × between subjects online experiment with N=410 participants. We developed novel use case which users go virtual hunt are tasked...

An increasing number of interactive visualization tools stress the integration with computational software like MATLAB and R to access a variety proven algorithms. In many cases, however, algorithms are used as black boxes that run completion in isolation which contradicts needs data exploration. This paper structures, formalizes, discusses possibilities enable user involvement ongoing computations. Based on structured characterization regarding intermediate feedback control, main...

Abstract Multivariate networks are made up of nodes and their relationships (links), but also data about those links as attributes. Most real‐world associated with several attributes, many analysis tasks depend on analyzing both, Visualization multivariate networks, however, is challenging, especially when both the topology network attributes need to be considered concurrently. In this state‐of‐the‐art report, we analyze current practices classify techniques along four axes: layouts, view...

Abstract There is fast‐growing literature on provenance‐related research, covering aspects such as its theoretical framework, use cases, and techniques for capturing, visualizing, analyzing provenance data. As a result, there an increasing need to identify taxonomize the existing scholarship. Such organization of research landscape will provide complete picture current state inquiry knowledge gaps or possible avenues further investigation. In this STAR, we aim produce comprehensive survey...

A common strategy in Multi-Criteria Decision Making (MCDM) is to rank alternative solutions by weighted summary scores. Weights, however, are often abstract the decision maker and can only be set vague intuition. While previous work supports a point-wise exploration of weight spaces, we argue that MCDM benefit from regional global visual analysis spaces. Our main contribution <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">WeightLifter</i> ,...

The realm of Artificial Intelligence (AI)'s impact on our lives is far reaching – with AI systems proliferating high-stakes domains such as healthcare, finance, mobility, law, etc., these must be able to explain their decision diverse end-users comprehensibly. Yet the discourse Explainable (XAI) has been predominantly focused algorithm-centered approaches, suffering from gaps in meeting user needs and exacerbating issues algorithmic opacity. To address issues, researchers have called for...

When analyzing multidimensional, quantitative data, the comparison of two or more groups dimensions is a common task. Typical sources such data are experiments in biology, physics engineering, which conducted different configurations and use replicates to ensure statistically significant results. One way analyze this filter it using statistical methods then run clustering algorithms group similar values. The results can be visualized heat maps, show differences between as changes color....

Evaluating, comparing, and interpreting related pieces of information are tasks that commonly performed during visual data analysis in many kinds information-intensive work. Synchronized highlighting elements is a well-known technique used to assist this task. An alternative approach, which more invasive but also expressive linking line connections rendered between elements. In work, we present context-preserving links as new method for generating links. The specifically aims fulfill the...

Abstract Identification and characterization of cancer subtypes are important areas research that based on the integrated analysis multiple heterogeneous genomics datasets. Since there no tools supporting this process, much work is done using ad‐hoc scripts static plots, which inefficient limits visual exploration data. To address this, we have developed StratomeX, an integrative visualization tool allows investigators to explore relationships candidate across genomic data types such as gene...

Abstract The primary goal of visual data exploration tools is to enable the discovery new insights. To justify and reproduce insights, process needs be documented communicated. A common approach documenting presenting findings capture visualizations as images or videos. Images, however, are insufficient for telling story a discovery, they lack full provenance information context. Videos difficult produce edit, particularly due non‐linear nature exploratory process. Most importantly, neither...

Abstract The introduction of machine learning to small molecule research– an inherently multidisciplinary field in which chemists and data scientists combine their expertise collaborate - has been vital making screening processes more efficient. In recent years, numerous models that predict pharmacokinetic properties or bioactivity have published, these are used on a daily basis by make decisions prioritize ideas. emerging explainable artificial intelligence is opening up new possibilities...

Explainable Artificial Intelligence (XAI) enables (AI) to explain its decisions. This holds the promise of making AI more understandable users, improving interaction, and establishing an adequate level trust. We tested this claim in high-risk task AI-assisted mushroom hunting, where people had decide whether a was edible or poisonous. In between-subjects experiment, 328 visitors Austrian media art festival played tablet-based hunting game while walking through highly immersive artificial...

Answering questions about complex issues often requires analysts to take into account information contained in multiple interconnected datasets. A common strategy analyzing and visualizing large heterogeneous data is dividing it meaningful subsets. Interesting subsets can then be selected the associated relationships between visualized. However, neither extraction manipulation nor comparison of well supported by state-of-the-art techniques. In this paper we present Domino, a novel multiform...

A major challenge in data-driven biomedical research lies the collection and representation of data provenance information to ensure that findings are reproducibile. In order communicate reproduce multi-step analysis workflows executed on datasets contain for dozens or hundreds samples, it is crucial be able visualize graph at different levels aggregation. Most existing approaches based node-link diagrams, which do not scale complexity typical graphs. our proposed approach, we reduce using...

Classifiers are among the most widely used supervised machine learning algorithms. Many classification models exist, and choosing right one for a given task is difficult. During model selection debugging, data scientists need to assess classifiers' performances, evaluate their behavior over time, compare different models. Typically, this analysis based on single-number performance measures such as accuracy. A more detailed evaluation of classifiers possible by inspecting class errors. The...

A typical problem in Visual Analytics (VA) is that users are highly trained experts their application domains, but have mostly no experience using VA systems. Thus, often difficulties interpreting and working with visual representations. To overcome these problems, user assistance can be incorporated into systems to guide through the analysis while closing knowledge gaps. Different types of applied extend power VA, enhance user's experience, broaden audience for VA. Although different...

The goal of our work is to support experts in the process hypotheses generation concerning roles genes diseases. For a deeper understanding complex interdependencies between genes, it important bring gene expressions (measurements) into context with pathways. Pathways, which are models biological processes, available online databases. In these databases, large networks decomposed small sub-graphs for better manageability. This simplification results loss context, as pathways interconnected...

Large volumes of real-world data often exhibit inhomogeneities: vertically in the form correlated or independent dimensions and horizontally clustered scattered items. In essence, these inhomogeneities patterns that researchers are trying to find understand. Sophisticated statistical methods available reveal patterns, however, visualization their outcomes is mostly still performed a one-view-fits-all manner. contrast, our novel approach, VisBricks, acknowledges inhomogeneity need for...

As heterogeneous data from different sources are being increasingly linked, it becomes difficult for users to understand how the connected, identify what means suitable analyze a given set, or find out proceed analysis task. We target this challenge with new model-driven design process that effectively codesigns aspects of data, view, analytics, and tasks. achieve by using workflow task as trajectory through interactive views, analytical processes. The benefits session go well beyond pure...

Biological pathway maps are highly relevant tools for many tasks in molecular biology. They reduce the complexity of overall biological network by partitioning it into smaller manageable parts. While this reduction is their biggest strength, is, at same time, weakness. By removing what deemed not important primary function pathway, biologists lose ability to follow and understand cross-talks between pathways. Considering these however, critical analysis scenarios, such as judging effects...

Cluster analysis is widely used to discover patterns in multi-dimensional data. Clustered heatmaps are the standard technique for visualizing one-way and two-way clustering results. In clustered heatmaps, rows and/or columns reordered, resulting a representation that shows clusters as contiguous blocks. However, biclustering results, where can overlap, it not possible reorder matrix this way without duplicating columns. We present Furby, an interactive visualization analyzing Our...

Storing analytical provenance generates a knowledge base with large potential for recalling previous results and guiding users in future analyses. However, without extensive manual creation of meta information annotations by the users, search retrieval analysis states can become tedious. We present KnowledgePearls, solution efficient that are structured as graphs containing automatically recorded user interactions visualizations. As core component, we describe visual interface querying...