IEEE VIS Publication Dataset

The final product of an analyst's investigation using a visualization is often a report of the discovered knowledge, as well as the methods employed and reasoning behind the discovery. We believe that analysts may have difficulty keeping track of their knowledge discovery process and will require tools to assist in accurately recovering their reasoning. We first report on a study examining analysts' recall of their strategies and methods, demonstrating their lack of memory of the path of knowledge discovery. We then explore whether a tool visualizing the steps of the visual analysis can aid users in recalling their reasoning process. The results of our second study indicate that visualizations of interaction logs can serve as an effective memory aid, allowing analysts to recall additional details of their strategies and decisions.

Modern visualization methods are needed to cope with very high-dimensional data. Efficient visual analytical techniques are required to extract the information content in these data. The large number of possible projections for each method, which usually grow quadrat-ically or even exponentially with the number of dimensions, urges the necessity to employ automatic reduction techniques, automatic sorting or selecting the projections, based on their information-bearing content. Different quality measures have been successfully applied for several specified user tasks and established visualization techniques, like Scatterplots, Scatterplot Matrices or Parallel Coordinates. Many other popular visualization techniques exist, but due to the structural differences, the measures are not directly applicable to them and new approaches are needed. In this paper we propose new quality measures for three popular visualization methods: Radviz, Pixel-Oriented Displays and Table Lenses. Our experiments show that these measures efficiently guide the visual analysis task.

Finding patterns in temporal data is an important data analysis task in many domains. Static visualizations can help users easily see certain instances of patterns, but are not specially designed to support systematic analysis tasks, such as finding all instances of a pattern automatically. VizPattern is an interactive visual query environment that uses a comic strip metaphor to enable users to easily and quickly define and locate complex temporal patterns. Evaluations provide evidence that VizPattern is applicable in many domains, and that it enables a wide variety of users to answer questions about temporal data faster and with fewer errors than existing state-of-the-art visual analysis systems.

Optimization problems are typically addressed by purely automatic approaches. For multi-objective problems, however, a single best solution often does not exist. In this case, it is necessary to analyze trade-offs between many conflicting goals within a given application context. This poster describes an approach that tightly integrates automatic algorithms for multi-objective optimization and interactive multivariate visualizations. Ad-hoc selections support a flexible definition of input data for subsequent algorithms. These algorithms in turn represent their result as derived data attributes that can be assigned to visualizations or be used as a basis for further selections (e.g., to constrain the result set). This enables a guided search that still involves the knowledge of domain experts. We describe our approach in the context of multi-run simulation data from the application domain of car engine design.

We present an interactive visual analytics system for classification, iVisClassifier, based on a supervised dimension reduction method, linear discriminant analysis (LDA). Given high-dimensional data and associated cluster labels, LDA gives their reduced dimensional representation, which provides a good overview about the cluster structure. Instead of a single two- or three-dimensional scatter plot, iVisClassifier fully interacts with all the reduced dimensions obtained by LDA through parallel coordinates and a scatter plot. Furthermore, it significantly improves the interactivity and interpretability of LDA. LDA enables users to understand each of the reduced dimensions and how they influence the data by reconstructing the basis vector into the original data domain. By using heat maps, iVisClassifier gives an overview about the cluster relationship in terms of pairwise distances between cluster centroids both in the original space and in the reduced dimensional space. Equipped with these functionalities, iVisClassifier supports users' classification tasks in an efficient way. Using several facial image data, we show how the above analysis is performed.

We present a unified framework for data processing, mining and interactive visualization of large-scale neuroanatomical databases. The input data is assumed to lie in a specific atlas space, or simply exist as a separate collection. Users can specify their own atlas for comparative analyses. The original data exist as MRI images in standard formats. It is uploaded to a remote server and processed offline by a parallelized pipeline workflow. This workflow transforms the data to represent it as both volumetric and triangular mesh cortical surfaces. We use multiresolution representations to scale complexity to data storage availability as well as graphical processing performance. Our workflow implements predefined metrics for clustering and classification, and data projection schemes to aid in visualization. Additionally the system provides a visual query interface for performing selection requests based on user-defined search criteria.

Visual exploration and analysis is a process of discovering and dissecting the abundant and complex attribute relationships that pervade multidimensional data. Recent research has identified and characterized patterns of multiple coordinated views, such as cross-filtered views, in which rapid sequences of simple interactions can be used to express queries on subsets of attribute values. In visualizations designed around these patterns, for the most part, distinct views serve to visually isolate each attribute from the others. Although the brush-and-click simplicity of visual isolation facilitates discovery of many-to-many relationships between attributes, dissecting these relationships into more fine-grained one-to-many relationships is interactively tedious and, worse, visually fragmented over prolonged sequences of queries. This paper describes: (1) a method for interactively dissecting multidimensional data by iteratively slicing and manipulating a multigraph representation of data values and value co-occurrences; and (2) design strategies for extending the construction of coordinated multiple view interfaces for dissection as well as discovery of attribute relationships in multidimensional data sets. Using examples from different domains, we describe how attribute relationship graphs can be combined with cross-filtered views, modularized for reuse across designs, and integrated into broader visual analysis tools. The exploratory and analytic utility of these examples suggests that an attribute relationship graph would be a useful addition to a wide variety of visual analysis tools.

Diagnosing faults in an operational computer network is a frustrating, time-consuming exercise. Despite advances, automatic diagnostic tools are far from perfect: they occasionally miss the true culprit and are mostly only good at narrowing down the search to a few potential culprits. This uncertainty and the inability to extract useful sense from tool output renders most tools not usable to administrators. To bridge this gap, we present NetClinic, a visual analytics system that couples interactive visualization with an automated diagnostic tool for enterprise networks. It enables administrators to verify the output of the automatic analysis at different levels of detail and to move seamlessly across levels while retaining appropriate context. A qualitative user study shows that NetClinic users can accurately identify the culprit, even when it is not present in the suggestions made by the automated component. We also find that supporting a variety of sensemaking strategies is a key to the success of systems that enhance automated diagnosis.

We propose a set of techniques that support visual interpretation of trajectory clusters by transforming absolute time references into relative positions within temporal cycles or with respect to the starting and/or ending times of the trajectories. We demonstrate the work of the approach on a real data set about individual movement over one year.

Complex combinations of coordinated multiple views are increasingly used to design tools for highly interactive visual exploration and analysis of multidimensional data. While complex coordination patterns provide substantial utility through expressive querying, they also exhibit usability problems for users when learning required interaction sequences, recalling past queries, and interpreting visual states. As visual analysis tools grow more sophisticated, there is a growing need to make them more understandable as well. Our long-term goal is to exploit natural language familiarity and literacy to directly facilitate individual and collaborative use of visual analysis tools. In this poster, we present work in progress on an automatically generated query-to-question user interface to translate interactive states during visual analysis into an accompanying visual log of formatted text. Our effort currently focuses on a symmetric and thus relatively simple coordination pattern: cross-filtered views. We describe our current thinking about query-to-question translation in a typical cross-filtered visualization of movies, people, and genres in the Internet Movie Database.

We present a custom visual analytics system developed in conjunction with the test and evaluation community of the US Army. We designed and implemented a visual programming environment for configuring a variety of interactive visual analysis capabilities. Our abstraction of the visualization process is based on insights gained from interviews conducted with expert users. We show that this model allowed analysts to implement multiple visual analysis capabilities for network performance, anomalous sensor activity, and engagement results. Long-term interaction with expert users led to development of several custom visual analysis techniques. We have conducted training sessions with expert users, and are working to evaluate the success of our work based on performance metrics captured in a semi-automated fashion during these training sessions. We have also integrated collaborative analysis features such as annotations and shared content.

Wikipedia has been built to gather encyclopedic knowledge using a collaborative social process that has proved its effectiveness. However, the workload required for raising the quality and increasing the coverage of Wikipedia is exhausting the community. Based on several participatory design sessions with active Wikipedia contributors (a.k.a. Wikipedians), we have collected a set of measures related to Wikipedia activity that, if available and visualized effectively, could spare a lot of monitoring time to these Wikipedians, allowing them to focus on quality and coverage of Wikipedia instead of spending their time navigating heavily to track vandals and copyright infringements. However, most of these measures cannot be computed on the fly using the available Wikipedia API. Therefore, we have designed an open architecture called WikiReactive to compute incrementally and maintain several aggregated measures on the French Wikipedia. This aggregated data is available as a Web Service and can be used to overlay information on Wikipedia articles through Wikipedia Skins or for new services for Wikipedians or people studying Wikipedia. This article describes the architecture, its performance and some of its uses.

These current studies explored the impact of individual differences in personality factors on interface interaction and learning performance behaviors in both an interactive visualization and a menu-driven web table in two studies. Participants were administered 3 psychometric measures designed to assess Locus of Control, Extraversion, and Neuroticism. Participants were then asked to complete multiple procedural learning tasks in each interface. Results demonstrated that all three measures predicted completion times. Additionally, results analyses demonstrated personality factors also predicted the number of insights participants reported while completing the tasks in each interface. We discuss how these findings advance our ongoing research in the Personal Equation of Interaction.

During the last decades, electronic textual information has become the world's largest and most important information source. Daily newspapers, books, scientific and governmental publications, blogs and private messages have grown into a wellspring of endless information and knowledge. Since neither existing nor new information can be read in its entirety, we rely increasingly on computers to extract and visualize meaningful or interesting topics and documents from this huge information reservoir. In this paper, we extend, improve and combine existing individual approaches into an overall framework that supports topologi-cal analysis of high dimensional document point clouds given by the well-known tf-idf document-term weighting method. We show that traditional distance-based approaches fail in very high dimensional spaces, and we describe an improved two-stage method for topology-based projections from the original high dimensional information space to both two dimensional (2-D) and three dimensional (3-D) visualizations. To demonstrate the accuracy and usability of this framework, we compare it to methods introduced recently and apply it to complex document and patent collections.

Text visualization becomes an increasingly more important research topic as the need to understand massive-scale textual information is proven to be imperative for many people and businesses. However, it is still very challenging to design effective visual metaphors to represent large corpora of text due to the unstructured and high-dimensional nature of text. In this paper, we propose a data model that can be used to represent most of the text corpora. Such a data model contains four basic types of facets: time, category, content (unstructured), and structured facet. To understand the corpus with such a data model, we develop a hybrid visualization by combining the trend graph with tag-clouds. We encode the four types of data facets with four separate visual dimensions. To help people discover evolutionary and correlation patterns, we also develop several visual interaction methods that allow people to interactively analyze text by one or more facets. Finally, we present two case studies to demonstrate the effectiveness of our solution in support of multi-faceted visual analysis of text corpora.

The detection of previously unknown, frequently occurring patterns in time series, often called motifs, has been recognized as an important task. To find these motifs, we use an advanced temporal data mining algorithm. Since our algorithm usually finds hundreds of motifs, we need to analyze and access the discovered motifs. For this purpose, we introduce three novel visual analytics methods: (1) motif layout, using colored rectangles for visualizing the occurrences and hierarchical relationships of motifs in a multivariate time series, (2) motif distortion, for enlarging or shrinking motifs as appropriate for easy analysis and (3) motif merging, to combine a number of identical adjacent motif instances without cluttering the display. We have applied and evaluated our methods using two real-world data sets: data center cooling and oil well production.

In risk assessment applications well informed decisions are made based on huge amounts of multi-dimensional data. In many domains not only the risk of a wrong decision, but in particular the trade-off between the costs of possible decisions are of utmost importance. In this paper we describe a framework tightly integrating interactive visual exploration with machine learning to support the decision making process. The proposed approach uses a series of interactive 2D visualizations of numeric and ordinal data combined with visualization of classification models. These series of visual elements are further linked to the classifier's performance visualized using an interactive performance curve. An interactive decision point on the performance curve allows the decision maker to steer the classification model and instantly identify the critical, cost changing data elements, in the various linked visualizations. The critical data elements are represented as images in order to trigger associations related to the knowledge of the expert. In this context the data visualization and classification results are not only linked together, but are also linked back to the classification model. Such a visual analytics framework allows the user to interactively explore the costs of his decisions for different settings of the model and accordingly use the most suitable classification model and make more informed and reliable decisions. A case study on data from the Forensic Psychiatry domain reveals the usefulness of the suggested approach.

The massive amount of financial time series data that originates from the stock market generates large amounts of complex data of high interest. However, adequate solutions that can effectively handle the information in order to gain insight and to understand the market mechanisms are rare. In this paper, we present two techniques and applications that enable the user to interactively analyze large amounts of time series data in real-time in order to get insight into the development of assets, market sectors, countries, and the financial market as a whole. The first technique allows users to quickly analyze combinations of single assets, market sectors as well as countries, compare them to each other, and to visually discover the periods of time where market sectors and countries get into turbulence. The second application clusters a selection of large amounts of financial time series data according to their similarity, and analyzes the distribution of the assets among market sectors. This allows users to identify the characteristic graphs which are representative for the development of a particular market sector, and also to identify the assets which behave considerably differently compared to other assets in the same sector. Both applications allow the user to perform investigative exploration techniques and interactive visual analysis in real-time.

We present a tool that is specifically designed to support a writer in revising a draft-version of a document. In addition to showing which paragraphs and sentences are difficult to read and understand, we assist the reader in understanding why this is the case. This requires features that are expressive predictors of readability, and are also semantically understandable. In the first part of the paper, we therefore discuss a semi-automatic feature selection approach that is used to choose appropriate measures from a collection of 141 candidate readability features. In the second part, we present the visual analysis tool VisRA, which allows the user to analyze the feature values across the text and within single sentences. The user can choose different visual representations accounting for differences in the size of the documents and the availability of information about the physical and logical layout of the documents. We put special emphasis on providing as much transparency as possible to ensure that the user can purposefully improve the readability of a sentence. Several case-studies are presented that show the wide range of applicability of our tool.

This paper presents an interactive interface synchronized with a simulation framework for exploring complex scenarios. This interface exploits visual analysis for facilitating the understanding of complex situation by human users.