In order to enhance analysis of synthetic health data of the IEEE vast challenge 2010, we introduce an interactive InfoVis tool called FilooT designed as a part of the Interactive Multi- genomic Analysis System (IMAS) project. In this paper we described different interactive views of FilooT: The tabular view for exploring and comparing genetic sequences, the matrix view for sorting sequences according to the values of different characteristics, P-value View for finding the most important mutations across a family of sequences, the Graph View for finding related sequences and the Group View to group them for further investigation. We followed the Nested Process Model framework throughout the design process and the evaluation. To understand the tool’s design capabilities for target domain analysts, we conducted a User Experience scenario-based study followed by an informal interview. The findings indicated how analysts employ each of the visualization and interaction designs in their Bioinformatics’ task-analysis process. The critical analysis of the results inspired design informing suggestions.

In recent years, real-time Magnetic Resonance Imaging (RT-MRI) has been used to acquire vocal tract data to support articulatory studies. The large amount of images resulting from these acquisitions needs to be processed and the resulting data analysed to extract articulatory features. This is often performed by linguists and phoneticists and requires not only tools providing a high level exploration of the data, to gather insight over the different aspects of speech, but also a set of features to compare different vocal tract configurations in static and dynamic scenarios. In order to make the data available in a faster and systematic fashion, without the continuous direct involvement of image processing specialists, a framework is being developed to bridge the gap between the more technical aspects of raw data and the higher level analysis required by speech researchers. In its current state it already includes segmentation of the vocal tract, allows users to explore the different aspects of the acquired data using coordinated views, and provides support for vocal tract configuration comparison. Beyond the traditional method of visual comparison of vocal tract profiles, a regional quantitative method is proposed, considering relevant anatomical features, and analysis supported by an abstract representation of the data both for static and dynamic analysis.

In this work, we present VAFLE, an interactive network security visualization prototype for the analysis of firewall log events. Keeping it simple yet effective for analysts, we provide multiple coordinated interactive visualizations augmented with clustering capabilities customized for the forensic analysis of network traffic. We evaluate the usefulness of the prototype in a use case with network traffic datasets from previous VAST Challenges, illustrating its effectiveness at promoting fast and well-informed decisions. We explain how a security analyst may spot suspicious traffic using VAFLE. We further assess its usefulness through a qualitative evaluation involving network security experts, whose feedback is reported and discussed.

The need to scale visualization of cyber (IP-space) data sets and analytic results as well as to support a variety of data sources and missions have proved challenging requirements for the development of a cyber common operating picture. Typical methods of visualizing IP-space data require unreliable domain conversions such as IP geolocation, network topology that is difficult to discover, or data sets that can only display one at a time. In this work, we introduce a generalized version of hierarchical network maps called configurable IP-space maps that can simultaneously visualize multiple layers of IP-based data at global scale. IP-space maps allow users to interactively explore the cyber domain from multiple perspectives. A web-based implementation of the concept is described, highlighting a novel repurposing of existing geospatial mapping tools for the cyber domain. Benefits of the configurable IP-space map concept to cyber data set analysis using spatial statistics are discussed. IP-space map structure is found to have a strong effect on data clustering behavior, hinting at the ability to automatically determine concentrations of network events within an organizational hierarchy.

Analysts need to keep track of their analytic findings, observations, ideas, and hypotheses throughout the analysis process. While some visual analytics tools support such note-taking needs, these notes are often represented as objects separate from the data and in a workspace separate from the data visualizations. Representing notes the same way as the data and integrating them with data visualizations can enable analysts to build a more cohesive picture of the analytical process. We created a note-taking functionality called CZNotes within the visual analytics tool CZSaw for analyzing unstructured text documents. CZNotes are designed to use the same model as the data and can thus be visualized in CZSaw's existing data views. We conducted a preliminary case study to observe the use of CZNotes and observed that CZNotes has the potential to support progressive analysis, to act as a shortcut to the data, and supports creation of new data relationships.

Monitoring and analysis of streaming data, such as social media, sensors, and news feeds, has become increasingly important for business and government. Volume and velocity of incoming data are key challenges. To effectively support monitoring and analysis, statistical and visual analytics techniques need to be seamlessly integrated; analytic techniques for a variety of data types (e.g. text, numerical) and scope (e.g. incremental, rolling-window, and global) must be properly accommodated; interaction and coordination among several visualizations must be supported in an efficient manner; and the system should support the use of different analytics techniques in a pluggable and collaborative manner. Especially in web based environments, these requirements pose restrictions on the basic architecture for such systems. In this paper we report on our experience of building a reference web architecture for real-time visual analytics of streaming data, identify and discuss architectural patterns that address these challenges, and report on applying the reference architecture for real-time Twitter monitoring.

General Recognition Theory (GRT) is a multidimensional signal detection theory framework for capturing sources of perceptual and decisional dependence. The primary type of data for GRT models is an identification-confusion matrix derived in a complete factorial identification task. This confusion matrix plots the responses of study participant for a given signal. The responses may reveal that participants were unable to recognize the signal properly. Such violations of any type of independence in the GRT framework result in response patterns that reflect some form of correlation in the GRT space. While an individual confusion matrix is rather small and relatively easy to visualize, similar studies may be repeated a lot of times resulting in thousands if not millions of confusion matrices that have to be visualized. This paper describes methods that adapt the web-based D3 visualization framework combined with pre-processing tools for the raw data to identify ways of visualization methodologies that enable domain specialists to more easily interpret their data. As the D3 framework utilizes Javascript and scalable vector graphics (SVG) to generate the visualizations it can run readily within the web browser to directly enable deployment of the visualization algorithms by the domain specialists. Parallel coordinate plots and heat maps were developed for the confusion matrix data, and the results were shown to a GRT expert for an informal evaluation of their utility. There is a clear benefit to model interpretation from these visualizations when researchers need to interpret larger amounts of simulated data.

We will explore the ongoing challenges and the future of cybersecurity visualization from an enterprise perspective. Based on an interview study with five cybersecurity experts serving in various roles in private and government sectors, we will characterize today’s cybersecurity environments based on over ten factors such as data size and velocity, data sensitivity, and goal/role of visualization. We will then present the top-ten observations that make cybersecurity visualizations different from other types of visualization.

Visual analytics and interactive machine learning both try to leverage the complementary strengths of humans and machines to solve complex data exploitation tasks. These fields overlap most significantly when training is involved: the visualization or machine learning tool improves over time by exploiting observations of the human-computer interaction. This paper focuses on one aspect of the human-computer interaction that we call user driven sampling strategies. Unlike relevance feedback and active learning sampling strategies, where the computer selects which data to label at each iteration, we investigate situations where the user selects which data is to be labeled at each iteration. User driven sampling strategies can emerge in many visual analytics applications but they have not been fully developed in machine learning. User driven sampling strategies suggest new theoretical and practical research questions for both visualization science and machine learning. In this paper we identify and quantify the potential benefits of these strategies in a practical image analysis application. We find user driven sampling strategies can sometimes provide significant performance gains by steering tools towards local minimum that have lower error than tools trained with all of the data. We find these performance gains are particularly pronounced when the user is experienced with the tool and application domain.

The human brain is the most sophisticated image processing system known, capable of impressive feats of recognition and discrimination under challenging natural conditions. Reverse-engineering the brain might enable us to design artificial systems with the same capabilities. My laboratory uses a data-driven system identification approach to tackle this reverse-engineering problem. Our approach consists of four broad stages. First, we use functional MRI to measure brain activity while people watch movies. We divide these data into two parts, one used to fit models and one for testing model predictions. Second, we use a system identification framework based on multiple linearizing feature spaces to model activity measured at each point in the brain. Third, we inspect the most accurate models to understand how the brain represents structural and semantic information in the movies. Finally, we use the estimated models to decode brain activity, reconstructing the structural and semantic content in the movies. This framework could form the basis of practical new brain reading technologies, and can inform development of biologically-inspired computer vision systems.

We present a tangible approach for exploring and comparing multi-dimensional data points collaboratively by combining Sifteo Cubes with glyph visualizations. Various interaction techniques like touching, shaking, moving or rotating the displays support the user in the analysis. Context dependent glyph-like visualization techniques make best use of the available screen space and cube arrangements. As a first proof of concept we apply our approach to real multi-dimensional datasets and show with a coherent use case how our techniques can facilitate the exploration and comparison of data points. Finally, further research directions are shown when combining Sifteo Cubes with glyphs and additional context information provided by multi-touch tables.

Visualizing data on tablets is challenging due to the relatively small screen size and limited user interaction capabilities. Standard data visualization apps provide support for pinch-and-zoom and scrolling operations, but do not provide context for data that is off-screen. When exploring data on tablets, the user must be able to focus on a region of interest and quickly find interesting patterns in the data. We present visualization techniques that facilitate seamless interaction with the region of interest on a tablet using context-providing bar graphs and scatter plots. Through aggregation, fisheye-style, and overview+detail representations, we provide context to the users as they explore a region of interest. We evaluated the efficacy of our techniques with the standard, interactive bar graph and scatter plot applications on a tablet, and found that one of our bargraph visualizations - Fisheye-style Focus+Context visualization (BG2) resulted in the fewest errors, least frustration and took the least amount of time. Similarly, one of our scatter plot visualizations - User Driven Overview+Detail (SP3) - resulted in the fewest errors, least frustration and took the least amount of time. Overall, users preferred the context-providing techniques over traditional bar graphs and scatter plots, that include pinch-and-zoom and fling-based scrolling capabilities.

This paper presents our interactive multi-dimensional visualization and the concept, working prototype and design space of a two-handed, hybrid spatial user interface for desktop VR targeted at users where a minimally immersive, desktop VR system is appropriate. The user interface supports dual button balls (6DOF isotonic controllers with multiple buttons) which automatically switch between 6DOF mode (xyz + yaw,pitch,roll) and planar-3DOF mode (xy + yaw) upon contacting the desktop. The mode switch automatically switches a button ball’s visual representation between a 3D cursor and a mouse-like 2D cursor while also switching the available user interaction techniques (ITs) between 3D and 2D ITs. Further, the small form factor of the button ball allows the user to engage in 2D multi-touch or 3D gestures without releasing and re-acquiring the device. We call the device and hybrid interface the HyFinBall interface which is an abbreviation for ‘Hybrid Finger Ball.’ We describe the user interface (hardware and software), the design space, as well as preliminary results of a formal user study. This is done in the context of a rich, visual analytics interface containing coordinated views with 2D and 3D visualizations and interactions

There are many systems that provide visualizations for time-oriented data. Of those, few provide the means of finding patterns in time-series data in which rankings are also important. Fewer still have the fine granularity necessary to visually follow individual data points through time. We propose the Ranking Timeline, a novel visualization method for modestly-sized multivariate data sets that include the top ten rankings over time. The system includes two main visualization components: a ranking over time and a cluster analysis. The ranking visualization, loosely based on line plots, allows the user to track individual data points so as to facilitate comparisons within a given time frame. Glyphs represent additional attributes within the framework of the overall system. The user has control over many aspects of the visualization, including viewing a subset of the data and/or focusing on a desired time frame. The cluster analysis tool shows the relative importance of individual items in conjunction with a visualization showing the connection(s) to other, similar items, while maintaining the aforementioned glyphs and user interaction. The user controls the clustering according to a similarity threshold. The system has been implemented as a Web application, and has been tested with data showing the top ten actors/actresses from 1929-2010. The experiments have revealed patterns in the data heretofore not explored.

In many application areas, the key to successful data analysis is the integrated analysis of heterogeneous data. One example is the fi nancial domain, where time-dependent and highly frequent quantitative data (e.g., trading volume and price information) and textual data (e.g., economic and political news reports) need to be considered jointly. Data analysis tools need to support an integrated analysis, which allows studying the relationships between textual news documents and quantitative properties of the stock market price series. In this paper, we describe a workflow and tool that allows a flexible formation of hypotheses about text features and their combinations, which reflect quantitative phenomena observed in stock data. To support such an analysis, we combine the analysis steps of frequent quantitative and text-oriented data using an existing a-priori method. First, based on heuristics we extract interesting intervals and patterns in large time series data. The visual analysis supports the analyst in exploring parameter combinations and their results. The identified time series patterns are then input for the second analysis step, in which all identified intervals of interest are analyzed for frequent patterns co-occurring with financial news. An a-priori method supports the discovery of such sequential temporal patterns. Then, various text features like the degree of sentence nesting, noun phrase complexity, the vocabulary richness, etc. are extracted from the news to obtain meta patterns. Meta patterns are defined by a specific combination of text features which significantly differ from the text features of the remaining news data. Our approach combines a portfolio of visualization and analysis techniques, including time-, cluster- and sequence visualization and analysis functionality. We provide two case studies, showing the effectiveness of our combined quantitative and textual analysis workflow. The workflow can also be generalized to other application domains such as data analysis of smart grids, cyber physical systems or the security of critical infrastructure, where the data consists of a combination of quantitative and textual time series data.

Increasing complexity and volume of data mandates integration between analytics and visualization. In this talk I am going to make a call for tighter integration of computational and visual analytics techniques, called Just-in-Time Visual Analytics, performed in real-time as users are exploring the data. I will argue that such close-coupling would bring about a novel user experience of computational techniques working on users’ behalf supporting a social, visual, and intelligent conversation with data, allowing them build faster qualitative mental models, speed up data and expertise discovery, and guide the discovery process. I will demonstrate just-in-time visual analytics through several demonstrations of systems I have built and discuss opportunities, and usability and architectural requirements necessary to build such systems that scale the user experience to the complexity of visual analytic discovery and to the skills of the ever-expanding user-base of visual analytics.

Often multidimensional data are visualized by splitting n-D data to a set of low dimensional data. While it is useful it destroys integrity of n-D data, and leads to a shallow understanding complex n-D data. To mitigate this difficulty a difficult perceptual task of assembling low-dimensional visualized pieces to the n-D record must be solved. An alternative way for deeper understanding of n-D data is visual representations of n-D data in low dimensions without data splitting. Methods of Parallel and Radial coordinates are such methods. Developing new methods of this type is a long standing task that this paper attempts to address by proposing Paired Coordinates that is a new type of n-D data visual representation and by generalizing Parallel and Radial coordinates as a General Line coordinates. The important novelty of the Paired Coordinates is that uses a single 2-D plot to represent n-D data as an oriented graph based on the idea of collocation of pairs of attributes. The advantage of the General Line Coordinates and Paired Coordinates is that they provide a common framework that includes Parallel and Radial coordinates and generate a large number of new visual representations of multidimensional data without dimension reduction.

Visualization researchers are more and more aware of the importance of evaluation, as it is not only a means of improving techniques and applications, but it can also produce evidence of measurable benefits that will encourage adoption. Yet, evaluating visualization applications or techniques, is not simple. This paper presents some issues that have to be considered while planning an evaluation and a brief description of the methods that the authors have been using to evaluate visualization techniques and applications concerning the users’ performance in their tasks when using the visualization, implying the understanding of the phenomenon. A list of guidelines considered as best practices to perform evaluations is also presented and some conclusions are drawn.

A five-star quality rating is one of the most widely used systems for evaluating items. However, it has two fundamental limitations: 1) the rating for one item cannot describe crucial information in detail; 2) the rating is not on an absolute scale for comparing to other items. Because of these limitations, users cannot make right decision. In this paper, we introduce our sophisticated approach to extract useful information from user reviews using collapsed dependencies and sentiment analysis. We propose an interactive word cloud that can show grammatical relationships among words, explore reviews efficiently, and display positivity or negativity on a sentence. In addition, we introduce visualization for comparing multiple word clouds and illustrate the usage through test cases.

Although shape perception is the main information channel for brain, it has been poor used by recent visualization techniques. The difficulties of its modeling are key obstacles for visualization theory and application. Known experimental estimates of shape perception capabilities have been made for low data dimension, and they were usually not connected with data structures. More applied approach for certain data structures detection by means of shape displays are considered by the example of analytical and experimental comparison of popular now Parallel Coordinates (PCs), i.e. 2D Cartesian displays of data vectors, with polar displays known as stars. Advantages of stars vs. PCs by Gestalt Laws are shown. About twice faster feature selection and classification with stars than PCs are showed by psychological experiments for hyper-tubes structures detection in data space with dimension up to 100-200 and its sub-spaces. This demonstrates great reserves of visualization enhancement in comparison with many recent techniques usually focused on few data attributes analysis.

X-ray Absorption Near Edge Structure (XANES) imaging, an advanced absorption spectroscopy technique, at the Transmission X-ray Microscopy (TXM) Beamline X8C of NSLS enables high-resolution chemical mapping (a.k.a. chemical composition identification or chemical spectra fitting). Two-Dimensional (2D) chemical mapping has been successfully applied to study many functional materials to decide the percentages of chemical components at each pixel position of the material images. In chemical mapping, the attenuation coefficient spectrum of the material (sample) can be fitted with the weighted sum of standard spectra of individual chemical compositions, where the weights are the percentages to be calculated. In this paper, we first implemented and compared two fitting approaches: (i) a brute force enumeration method, and (ii) a constrained least square minimization algorithm proposed by us. Next, as 2D spectra fitting can be conducted pixel by pixel, so theoretically, both methods can be implemented in parallel. In order to demonstrate the feasibility of parallel computing in the chemical mapping problem and investigate how much efficiency improvement can be achieved, we used the second approach as an example and implemented a parallel version for a multi-core computer cluster. Finally we used a novel way to visualize the calculated chemical compositions, by which domain scientists could grasp the percentage difference easily without looking into the real data.

A significant task within data mining is to identify data models of interest. While facilitating the exploration tasks, most visualization systems do not make use of all the data models that are generated during the exploration. In this paper, we introduce a system that allows the user to gain insights from the data space progressively by forming data models and consolidating the generated models on the fly. Each model can be a a computationally extracted or user-defined subset that contains a certain degree of interest and might lead to some discoveries. When the user generates more and more data models, the degree of interest of some portion of some models will either grow (indicating higher occurrence) or will fluctuate or decrease (corresponding to lower occurrence). Our system maintains a collection of such models and accumulates the interestingness of each model into a consolidated model. In order to consolidate the models, the system summarizes the associations between the models in the collection and identifies support (models reinforce each other), complementary (models complement each other), and overlap of the models. The accumulated interestingness keeps track of historical exploration and helps the user summarize their findings which can lead to new discoveries. This mechanism for integrating results from multiple models can be applied to a wide range of decision support systems. We demonstrate our system in a case study involving the financial status of US companies.

A web-based application has been developed for system administrators to visualize metadata of Lustre file system. The application is sub-divided into two visualizations to monitor different attributes about the users of Kraken, a University of Tennessee High Performance Computing (HPC) resource in Oak Ridge National Laboratory (ORNL). In this presentation, the development processes using D3 and Highcharts Javascript libraries are shown to demonstrate the capabilities and difficulties of both libraries. Moreover, in the difficulties part, the focus is on profiling the execution time when visualizing different sizes of data sets.

Spatial partitions of an information space are frequently used for data visualization. Weighted Voronoi diagrams are among the most popular ways of dividing a space into partitions. However, the problem of computing such a partition efficiently can be challenging. For example, a natural objective is to select the weights so as to force each Voronoi region to take on a pre-defined area, which might represent the relevance or market share of an informational object. In this paper, we present an easy and fast algorithm to compute these weights of the Voronoi diagrams. Unlike previous approaches whose convergence properties are not well-understood, we give a formulation to the problem based on convex optimization with excellent performance guarantees in theory and practice. We also show how our technique can be used to control the shape of these partitions. More specifically we show how to convert undesirable skinny and long regions into fat regions while maintaining the areas of the partitions. As an application, we use these to visualize the amount of website traffic for the top 101 websites.

Probabilistic modeling and visualization of crater shape-maturity relationships is explored in context to the remote sensing data acquired from Apollo, Clementine and Lunar Reconnaissance Orbiter lunar missions. Unlike any earlier attempt in understanding relationships between lunar crater features (depth and diameter), relative age of crater formation (Pre-Nectarian to Copernican) and optical maturity of the lunar surface (OMAT values), the joint probability of these variables is modeled in this work. The proposed model is strongly dependent on data density and is not based on deterministic equations as in earlier works. Once developed,a joint probability model can accommodate additional factors through conditional probability weights in a Bayesian network architecture. It is expected that probabilistic modeling will facilitate visualization of relationships between experimental variables and eventually help gain additional insight into lunar cratering mechanisms and linkages between crater morphology, spectral properties and crater degradation mechanisms. The described simple Bayesian network in this work is by no means complete, but illustrates the potential of the proposed novel method in the context of ever-increasing high resolution lunar data in recent times.

This paper produces a guideline for the development of information visualization tools for social networks using it for this a case study on the psychological behavior of people in social networks. At the end of the paper are presented results of evaluations of usability and performance, compared to other visualization tools for social networking.

Several presentation techniques have been created for visualization of data with more than three variables. Packages have been written, each of which implements a subset of these techniques. However, these packages generally fail to provide all the features needed by the user during the visualization process. Further, packages generally limit support for presentation techniques to a few techniques. A new package called Petrichor accommodates all necessary and useful features together in one system. Any presentation technique may be added easily through an extensible plugin system. Features are supported by a user interface that allows easy interaction with data. Annotations allow users to mark up visualizations and share information with others. By providing a hyperdimensional graphics package that easily accommodates presentation techniques and includes a complete set of features, including those that are rarely or never supported elsewhere, the user is provided with a tool that facilitates improved interaction with multivariate data to extract and disseminate information.

Some people suggest that imaging is a mature field, but nothing could be further from the truth. In fact, we are entering into the next phase of innovation in which a convergence of technologies is enabling the creation of an endless variety of imaging systems based on the tight integration of algorithms, computation, and sensor design. This new field, which we call integrated imaging, is evolving out of classical imaging modalities into a wide array of new applications. Integrated imaging systems will drive both scientific exploration and consumer products by blending novel and often counter-intuitive sensor design with algorithms that exploit the availability of enormous quantities of data and computation. This talk presents some examples of state-of-the-art integrated imaging systems based on computed tomography (CT), transmission electron microscopy (STEM), synchrotron beam imaging, optical sensing, and scanning electron microscopy (SEM). For each of these examples, we also explore their use and potential impact in applications ranging from healthcare to jet engine design. We conclude with some speculation on where integrated imaging might be going; where it might have greatest impact; and what will be the greatest challenges ahead.

Today's tracking devices allow high spatial and temporal resolutions and due to their decreasing size also an ever increasing number of application scenarios. However, understanding motion over time is quite difficult as soon as the resulting trajectories are getting complex. Simply plotting the data may obscure important patterns since trajectories over long time periods often include many revisits of the same place which creates a high degree of over-plotting. Furthermore, important details are often hidden due to a combination of large-scale transitions with local and small-scale movement patterns. We present a visualization and abstraction technique for such complex motion data. By analyzing the motion patterns and displaying them with visual abstraction techniques a synergy of aggregation and simplification is reached. The capabilities of the method are shown in real-world applications for tracked animals and discussed with experts from biology. Our proposed abstraction techniques reduce visual clutter and help analysts to understand the movement patterns that are hidden in raw spatiotemporal data.

As visualization is applied to larger data sets residing in more diverse hardware environments, visualization frameworks need to adapt. Rendering techniques are currently a major limiter since they tend to be built around central processing with all of the geometric data present. This is not a fundamental requirement of information visualization. This paper presents Abstract Rendering, a technique for eliminating the centralization requirement while preserving some forms of interactivity. Abstract Rendering is based on the observation that pixels are fundamentally bins, and that rendering is essentially a binnig process on a lattice of bins. By providing a more flexible binning process, the majority of rendering can be done with the geometric information stored out-of-core. Only the bin representations need to reside in memory. This approach enables rendering on large datasets without requiring large amounts of working memory, novel and useful control over image composition, a direct means of distributing the rendering task across processes and high-performance interaction techniques on large datasets. This paper introduces Abstract Rendering, provides an overview of an implementation and discuss how it has been applied to some large data visualization problems.

Understanding of vector fields is important in many science and engineering domains. Traditionally, glyphs have been used to represent vector data as cones, arrows, ellipsoids and other geometric shapes. However, these glyphs have drawbacks of being view dependent, orientation ambiguous, and sometimes-even requiring specific geometry resolution. We propose a straightforward new method of procedural dipole texturing of glyph shapes, which overcomes these drawbacks and could even enhance existing methods. We demonstrate our method with an interactive application (GlyphSea), which incorporates additional features such as screen space ambient occlusion, displacements, lattice, halos and other contextual visual cues. We also discuss the results and informal feedback from scientists on insights gained by exploring time varying vector datasets in astrophysics and seismology.

Capillary electrophoresis and similar techniques which use an electrified contracting-flow interface (gradient elution moving boundary electrophoresis, electrophoretic exclusion, for examples) are widely used and have several applications, but the detailed flow dynamics and local electric field effects within this zone have only recently been quantitatively investigated. The motivating force behind this work is establishing particle flow based visualization tools enabling advances for arbitrary interfacial designs beyond this traditional flow/electric field interface. These tools work with pre-computed 2-dimensional fundamental interacting fields which govern particle and(or) fluid flow and can now be obtained from various computational fluid dynamics (CFD) software packages. The particle-flow visualization calculations implemented in the tool and are built upon a solid foundation in fluid dynamics. The module developed in here provides a simulated video particle observation tool which generates a fast check for legitimacy. Further, estimating the accuracy and precision of full 2-D and 3-D simulation is notoriously difficult and a centerline estimation is used to quickly and easily quantitate behaviors in support of decision points. This tool and the recent quantitative assessment of particle behavior within the interfacial area have set the stage for new designs which can emphasize advantageous behaviors not offered by the traditional configuration.

Streamline similarity comparison has become an active research topic recently. In this paper, we present a novel streamline similarity comparison method inspired by the bag-of-features idea from computer vision. Our approach computes a feature vector, spatially sensitive bag-of-features, for each streamline as its signature. This feature vector not only encodes the statistical distribution of combined features (e.g., curvature and torsion), it also contains the information on the spatial relationship among different features. This allows us to measure the similarity between two streamlines in an efficient and accurate way: the similarity between two streamlines is defined as the weighted Manhattan distance between their feature vectors. Compared with previous distribution based streamline similarity metrics, our method is easier to understand and implement, yet producing even better results. We demonstrate the utility of our approach by considering two common tasks in flow field exploration: streamline similarity query and streamline clustering.