@conference {DOI336, title = {Evaluating Differences in Insights from Interactive Dimensionality Reduction Visualizations through Complexity and Vocabulary}, booktitle = {International Conference on Information Visualization Theory and Applications (IVAPP)}, year = {2023}, month = {02/2023}, pages = {8 pages}, author = {Mia Taylor and Lata Kodali and House, Leanna and North, Chris} } @conference {asee_peer_41168, title = {Andromeda in the Classroom: Collaborative Data Analysis for 8th Grade Engineering Design}, booktitle = {2022 ASEE Annual Conference \& Exposition}, year = {2022}, note = {https://peer.asee.org/41168}, month = {08/2022}, publisher = {ASEE Conferences}, organization = {ASEE Conferences}, address = {Minneapolis, MN}, author = {Mia Taylor and Danny Mathieson and House, Leanna and North, Chris} } @article {DOI329, title = {Bridging cognitive gaps between user and model in interactive dimension reduction}, journal = {Visual Informatics}, volume = {53}, year = {2021}, pages = {13-25}, author = {Ming Wang and Wenskovitch, John and House, Leanna and Nicholas Polys and North, Chris} } @article {DOWLING201949, title = {Interactive Visual Analytics for Sensemaking with Big Text}, journal = {Big Data Research}, volume = {16}, year = {2019}, month = {July/2019}, pages = {49 - 58}, abstract = {Analysts face many steep challenges when performing sensemaking tasks on collections of textual information larger than can be reasonably analyzed without computational assistance. To scale up such sensemaking tasks, new methods are needed to interactively integrate human cognitive sensemaking activity with machine learning. Towards that goal, we offer a human-in-the-loop computational model that mirrors the human sensemaking process, and consists of foraging and synthesis sub-processes. We model the synthesis loop as an interactive spatial projection and the foraging loop as an interactive relevance ranking combined with topic modeling. We combine these two components of the sensemaking process using semantic interaction such that the human{\textquoteright}s spatial synthesis actions are transformed into automated foraging and synthesis of new relevant information. Ultimately, the model{\textquoteright}s ability to forage as a result of the analyst{\textquoteright}s synthesis activities makes interacting with big text data easier and more efficient, thereby facilitating analysts{\textquoteright} sensemaking ability. We discuss the interaction design and theory behind our interactive sensemaking model. The model is embodied in a novel visual analytics prototype called Cosmos in which analysts synthesize structure within the larger corpus by directly interacting with a reduced-dimensionality space to express relationships on a subset of data. We then demonstrate how Cosmos supports sensemaking tasks with a realistic scenario that investigates the affect of natural disasters in Adelaide, Australia in September 2016 using a database of over 30,000 news articles.}, keywords = {Big data, interactive visual analytics, Semantic interaction, text analytics, Topic modeling, visualization}, issn = {2214-5796}, doi = {https://doi.org/10.1016/j.bdr.2019.04.003}, url = {http://www.sciencedirect.com/science/article/pii/S2214579618302995}, author = {Michelle Dowling and Nathan Wycoff and Brian Mayer and Wenskovitch, John and Leman, Scotland and House, Leanna and Nicholas Polys and North, Chris and Peter Hauck} } @proceedings {kodali2019uncertainty, title = {Uncertainty in Interactive WMDS Visualizations}, journal = {2019 Symposium on Visualization in Data Science Posters}, year = {2019}, address = {Vancouver, BC, Canada}, abstract = {Visualizations are useful when learning from high-dimensional data. However, visualizations can be misleading when they do not incorporate measures of uncertainty; e.g., uncertainty from the data or the dimension reduction algorithm used to create the visual display. In our work, we extend a framework called Bayesian Visual Analytics (BaVA) on a dimension reduction algorithm, Weighted Multidimensional Scaling (WMDS), to incorporate uncertainty as analysts explore data visually. BaVA-WMDS visualizations are interactive, and possible interactions include manipulating variable weights and/or the coordinates of the two-dimensional projection. Uncertainty exists in these visualizations on the variable weights, the user interactions, and the fit of WMDS. We quantify these uncertainties using Bayesian models exploring randomness in both coordinates and weights in a method we call Interactive Probabilistic WMDS (IP-WMDS). Specifically, we use posterior estimates to assess fit of WMDS, the range of motion of coordinates, as well as variability in weights. Visually, we display such uncertainty in the form of color and ellipses, and practically, these uncertainties reflect trust in fitting a dimension reduction algorithm. Our results show that these displays of uncertainty highlight different aspects of the visualization, which can help inform analysts.}, keywords = {poster}, author = {Lata Kodali and Wenskovitch, John and Nathan Wycoff and House, Leanna and North, Chris} } @article {DOI269, title = {Be the Data: Embodied Visual Analytics}, journal = {IEEE Transactions on Learning Technologies}, volume = {11}, year = {2018}, pages = {81-95}, doi = {10.1109/TLT.2017.2757481}, author = {Xin Chen and Self, Jessica Zeitz and House, Leanna and Wenskovitch, John and Sun, Maoyuan and Nathan Wycoff and Jane Robertson Evia and Leman, Scotland and North, Chris} } @conference {wenskovitch2018effect, title = {The Effect of Semantic Interaction on Foraging in Text Analysis}, booktitle = {2018 IEEE Conference on Visual Analytics Science and Technology (VAST)}, year = {2018}, abstract = {Completing text analysis tasks is a continuous sensemaking loop of foraging for information and incrementally synthesizing it into hypotheses. Past research has shown the advantages of using spatial workspaces as a means for synthesizing information through externalizing hypotheses and creating spatial schemas. However, spatializing the entirety of datasets becomes prohibitive as the number of documents available to the analysts grows, particularly when only a small subset are relevant to the task at hand. StarSPIRE is a visual analytics tool designed to explore collections of documents, leveraging users{\textquoteright} semantic interactions to steer (1) a synthesis model that aids in document layout, and (2) a foraging model to automatically retrieve new relevant information. In contrast to traditional keyword search foraging (KSF), "semantic interaction foraging" (SIF) occurs as a result of the user{\textquoteright}s synthesis actions. To quantify the value of semantic interaction foraging, we use StarSPIRE to evaluate its utility for an intelligence analysis sensemaking task. Semantic interaction foraging accounted for 26\% of useful documents found, and it also resulted in increased synthesis interactions and improved sensemaking task performance by users in comparison to only using keyword search.}, author = {Wenskovitch, John and Lauren Bradel and Michelle Dowling and House, Leanna and North, Chris} } @article {self2018observation, title = {Observation-Level and Parametric Interaction for High-Dimensional Data Analysis}, journal = {ACM Transactions on Interactive Intelligent Systems}, volume = {8}, year = {2018}, month = {07/2018}, doi = {10.1145/3158230}, author = {Self, Jessica Zeitz and Michelle Dowling and Wenskovitch, John and Ian Crandell and Ming Wang and House, Leanna and Leman, Scotland and North, Chris} } @conference {dowling2018sirius, title = {SIRIUS: Dual, Symmetric, Interactive Dimension Reductions}, booktitle = {2018 IEEE Conference on Visual Analytics Science and Technology (VAST)}, year = {2018}, month = {Oct}, abstract = {Much research has been done regarding how to visualize and interact with observations and attributes of high-dimensional data for exploratory data analysis. From the analyst{\textquoteright}s perceptual and cognitive perspective, current visualization approaches typically treat the observations of the high-dimensional dataset very differently from the attributes. Often, the attributes are treated as inputs (e.g., sliders), and observations as outputs (e.g., projection plots), thus emphasizing investigation of the observations. However, there are many cases in which analysts wish to investigate both the observations and the attributes of the dataset, suggesting a symmetry between how analysts think about attributes and observations. To address this, we define SIRIUS (Symmetric Interactive Representations In a Unified System), a symmetric, dual projection technique to support exploratory data analysis of high-dimensional data. We provide an example implementation of SIRIUS and demonstrate how this symmetry affords additional insights.}, keywords = {attribute projection, dimension reduction, exploratory data analysis, observation projection, Semantic interaction}, author = {Michelle Dowling and Wenskovitch, John and J.T. Fry and Leman, Scotland and House, Leanna and North, Chris} } @article {wenskovitch2018towards, title = {Towards a Systematic Combination of Dimension Reduction and Clustering in Visual Analytics}, journal = {IEEE Transactions on Visualization and Computer Graphics}, volume = {24}, number = {01}, year = {2018}, month = {01/2018}, pages = {131-141}, keywords = {Algorithm design and analysis, clustering, Clustering algorithms, Data visualization, Dimension reduction;algorithms, Manifolds, Partitioning algorithms, Visual Analytics, visualization}, issn = {1077-2626}, doi = {10.1109/TVCG.2017.2745258}, author = {Wenskovitch, John and Ian Crandell and Ramakrishnan, Naren and House, Leanna and Leman, Scotland and North, Chris} } @conference {DOI265, title = {Bringing Interactive Visual Analytics to the Classroom for Developing EDA Skills}, booktitle = {Proceedings of the 33rd Annual Consortium of Computing Sciences in Colleges (CCSC) Eastern Regional Conference}, year = {2017}, month = {10/2017}, pages = {10}, author = {Self, Jessica Zeitz and Self, Nathan and House, Leanna and Jane Robertson Evia and Leman, Scotland and North, Chris} } @conference {DOI251, title = {Be the Data: A New Approach for Immersive Analytics}, booktitle = {IEEE Virtual Reality 2016 Workshop on Immersive Analytics}, year = {2016}, month = {03/2016}, pages = {6}, author = {Xin Chen and Self, Jessica Zeitz and House, Leanna and North, Chris} } @conference {DOI245, title = {Be the Data: An Embodied Experience for Data Analytics}, booktitle = { 2016 Annual Meeting of the American Educational Research Association (AERA)}, year = {2016}, month = {04/2016}, pages = {20}, author = {Xin Chen and House, Leanna and Self, Jessica Zeitz and Leman, Scotland and Jane Robertson Evia and James Thomas Fry and North, Chris} } @conference {DOI248, title = {Be the Data: Social Meetings with Visual Analytics}, booktitle = {International Workshop on Visualization and Collaboration (VisualCol 2016)}, year = {2016}, month = {11/2016}, pages = {8}, author = {Xin Chen and Self, Jessica Zeitz and Sun, Maoyuan and House, Leanna and North, Chris} } @conference {DOI250, title = {Designing Usable Interactive Visual Analytics Tools for Dimension Reduction}, booktitle = {CHI 2016 Workshop on Human-Centered Machine Learning (HCML)}, year = {2016}, month = {05/2016}, pages = {7}, author = {Self, Jessica Zeitz and Hu, Xinran and House, Leanna and Leman, Scotland and North, Chris} } @article {DOI235, title = {Andromeda: Observation-Level and Parametric Interaction for Exploratory Data Analysis}, year = {2015}, institution = {Virginia Tech}, type = {Technical Report}, address = {Blacksburg}, abstract = {Exploring high-dimensional number of dimensions in datasets increases, it becomes harder to discover patterns and develop insights. Dimension reduction algorithms, such as multidimensional scaling, support data explorations by reducing datasets to two dimensions for visualization. Because these algorithms rely on underlying parameterizations, they may be tweaked to assess the data from multiple perspectives. Alas, tweaking can be difficult for users without a strong knowledge base of the underlying algorithms. We present Andromeda, an interactive visual analytics tool we developed to enable non-experts of statistical models to explore domain- specific, high-dimensional data. This application implements interactive weighted multidimensional scaling (WMDS) and allows for both parametric and observation- level interaction to provide in-depth data exploration. In this paper, we present the results of a controlled usability study assessing Andromeda. We focus on the comparison of parametric interaction, observation-level interaction and a combination of the two.}, author = {Self, Jessica Zeitz and House, Leanna and Leman, Scotland and North, Chris} } @conference {DOI242, title = {Big Text Visual Analytics in Sensemaking}, booktitle = {IEEE International Symposium on Big Data Visual Analytics}, year = {2015}, month = {09/2015}, pages = {8 pages}, author = {Lauren Bradel and Nathan Wycoff and House, Leanna and North, Chris} } @article {DOI244, title = {Bringing Interactive Visual Analytics to the Classroom for Developing EDA Skills}, year = {2015}, institution = {Virginia Tech}, type = {Technical Report}, address = {Blacksburg}, abstract = {This paper addresses the use of visual analytics in education for teaching what we call cognitive dimensionality (CD) and other EDA skills. We present the concept of CD to characterize students{\textquoteright} capacity for making dimensionally complex insights from data. Using this concept, we build a vocabulary and methodology to support a student{\textquoteright}s progression in terms of growth from low cognitive dimensionality (LCD) to high cognitive dimensionality (HCD). Crucially, students do not need high-level math skills to develop HCD. Rather, we use our own tool called Andromeda that enables human-computer interaction with a common, easy to interpret visualization method called Weighted Multidimensional Scaling (WMDS) to promote the idea of making high-dimensional insights. In this paper, we present Andromeda and report findings from a series of classroom assignments to 18 graduate students. These assignments progress from spreadsheet manipulations to statistical software such as R and finally to the use of Andromeda. In parallel with the assignments, we saw students{\textquoteright} CD begin low and improve.}, keywords = {dimension reduction, education, multidimensional scaling, multivariate analysis, Visual Analytics}, author = {Self, Jessica Zeitz and Self, Nathan and House, Leanna and Jane Robertson Evia and Leman, Scotland and North, Chris} } @article {DOI241, title = {Designing for Interactive Dimension Reduction Visual Analytics Tools to Explore High-Dimensional Data}, year = {2015}, institution = {Virginia Tech}, type = {Technical Report}, address = {Blacksburg}, abstract = {Exploring high-dimensional data is challenging. As the number of dimensions in datasets increases, the harder it becomes to discover patterns and develop insights. Dimension reduction algorithms, such as multidimensional scaling, support data explorations by reducing datasets to two dimensions for visualization. Because these algorithms rely on underlying parameterizations, they may be tweaked to assess the data from multiple perspectives. Alas, tweaking can be difficult for users without a strong knowledge base of the underlying algorithms. In this paper, we present principles for developing interactive visual analytic systems that enable users to tweak model parameters directly or indirectly so that they may explore high-dimensional data. To exemplify our principles, we introduce an application that implements interactive weighted multidimensional scaling (WMDS). Our application, Andromeda, allows for both parametric and object-level interaction to provide in-depth data exploration. In this paper, we describe the types of tasks and insights that users may gain with Andromeda. Also, the final version of Andromeda is the result of sequential improvements made to multiple designs that were critiqued by users. With each critique we uncovered design principles of effective, interactive, visual analytic tools. These design principles focus on three main areas: (1) layout, (2) semantically visualizing parameters, and (3) designing the communication between the interface and the algorithm.}, author = {Self, Jessica Zeitz and Hu, Xinran and House, Leanna and Leman, Scotland and North, Chris} } @article {DOI228, title = {Improving Students{\textquoteright} Cognitive Dimensionality through Education with Object-Level Interaction}, year = {2014}, institution = {Virginia Tech}, type = {Technical Report}, address = {Blacksburg}, abstract = {This paper addresses the use of visual analytics techniques in education to advance students{\textquoteright} cognitive dimensionality. Students naturally tend to characterize data in simplistic one dimensional terms using metrics such as mean, median, mode. Real- world data, however, is more complex and students need to learn to recognize and create high-dimensional arguments. Data exploration methods can encourage thinking beyond traditional one dimensional insights. In particular, visual analytics tools that afford object-level interaction (OLI) allow for generation of more complex insights, despite inexperience with multivariate data. With these tools, students{\textquoteright} insights are of higher complexity in terms of dimensionality and cardinality and built on more diverse interactions. We present the concept of cognitive dimensionality to characterize students{\textquoteright} capacity for dimensionally complex insights. Using this concept, we build a vocabulary and methodology to support a student{\textquoteright}s progression in terms of growth from low to high cognitive dimensionality. We report findings from a series of classroom assignments with increasingly complex analysis tools. These assignments progressed from spreadsheet manipulations to statistical software such as R and finally to an OLI application, Andromeda. Our findings suggest that students{\textquoteright} cognitive dimensionality can be improved and further research on the impact of visual analytics tools on education for cognitive dimensionality is warranted.}, keywords = {multivariate data analysis, object level interaction, Visual Analytics}, author = {Self, Jessica Zeitz and Self, Nathan and House, Leanna and Leman, Scotland and North, Chris} } @conference {DOI219, title = {StarSpire: Multi-Model Semantic Interaction for Text Analytics}, booktitle = {IEEE Conference on Visual Analytics Science and Technology (VAST)}, year = {2014}, pages = {1-10}, publisher = {IEEE}, organization = {IEEE}, address = {Paris, France}, abstract = {Semantic interaction offers an intuitive communication mechanism between human users and complex statistical models. By shielding the users from manipulating model parameters, they focus instead on directly manipulating the spatialization, thus remaining in their cognitive zone. However, this technique is not inherently scalable past hundreds of text documents. To remedy this, we present the concept of multi-model semantic interaction, where semantic interactions can be used to steer multiple models at multiple levels of data scale, enabling users to tackle larger data problems. We also present an updated visualization pipeline model for generalized multi-model semantic interaction. To demonstrate multi-model semantic interaction, we introduce StarSPIRE, a visual text analytics prototype that transforms user interactions on documents into both small-scale display layout updates as well as large-scale relevancy-based document selection.}, author = {Lauren Bradel and North, Chris and House, Leanna and Leman, Scotland} } @article {DOI10.1109/TVCG.2013.188, title = {Semantics of Directly Manipulating Spatializations}, journal = {IEEE Transactions on Visualization and Computer Graphics}, volume = {19}, year = {2013}, month = {12/2013}, pages = {2052 - 2059}, issn = {1077-2626}, doi = {10.1109/TVCG.2013.188}, author = {Hu, Xinran and Lauren Bradel and Maiti, Dipayan and House, Leanna and North, Chris and Leman, Scotland} } @article {DOI10.1371/journal.pone.0050474, title = {Visual to Parametric Interaction (V2PI)}, journal = {PLoS ONE}, volume = {8}, year = {2013}, month = {03/2013}, pages = {e50474}, abstract = {Typical data visualizations result from linear pipelines that start by characterizing data using a model or algorithm to reduce the dimension and summarize structure, and end by displaying the data in a reduced dimensional form. Sensemaking may take place at the end of the pipeline when users have an opportunity to observe, digest, and internalize any information displayed. However, some visualizations mask meaningful data structures when model or algorithm constraints (e.g., parameter specifications) contradict information in the data. Yet, due to the linearity of the pipeline, users do not have a natural means to adjust the displays. In this paper, we present a framework for creating dynamic data displays that rely on both mechanistic data summaries and expert judgement. The key is that we develop both the theory and methods of a new human-data interaction to which we refer as {\textquotedblleft} Visual to Parametric Interaction{\textquotedblright} (V2PI). With V2PI, the pipeline becomes bi-directional in that users are embedded in the pipeline; users learn from visualizations and the visualizations adjust to expert judgement. We demonstrate the utility of V2PI and a bi-directional pipeline with two examples.}, doi = {10.1371/journal.pone.0050474}, author = {Leman, Scotland and House, Leanna and Maiti, Dipayan and Endert, Alex and North, Chris} } @conference {6102449, title = {Observation-level interaction with statistical models for visual analytics}, booktitle = {Visual Analytics Science and Technology (VAST), 2011 IEEE Conference on}, year = {2011}, month = {oct.}, pages = {121 -130}, abstract = {In visual analytics, sensemaking is facilitated through interactive visual exploration of data. Throughout this dynamic process, users combine their domain knowledge with the dataset to create insight. Therefore, visual analytic tools exist that aid sensemaking by providing various interaction techniques that focus on allowing users to change the visual representation through adjusting parameters of the underlying statistical model. However, we postulate that the process of sensemaking is not focused on a series of parameter adjustments, but instead, a series of perceived connections and patterns within the data. Thus, how can models for visual analytic tools be designed, so that users can express their reasoning on observations (the data), instead of directly on the model or tunable parameters? Observation level (and thus $\#$x201C;observation $\#$x201D;) in this paper refers to the data points within a visualization. In this paper, we explore two possible observation-level interactions, namely exploratory and expressive, within the context of three statistical methods, Probabilistic Principal Component Analysis (PPCA), Multidimensional Scaling (MDS), and Generative Topographic Mapping (GTM). We discuss the importance of these two types of observation level interactions, in terms of how they occur within the sensemaking process. Further, we present use cases for GTM, MDS, and PPCA, illustrating how observation level interaction can be incorporated into visual analytic tools.}, keywords = {data analysis, data interactive visual exploration, data visualisation, exploratory interaction, expressive interaction, generative topographic mapping, multidimensional scaling, observation-level interaction, parameter adjustments, principal component analysis, probabilistic principal component analysis, probability, sensemaking process, statistical models, Visual Analytics}, doi = {10.1109/VAST.2011.6102449}, author = {Endert, Alex and Chao Han and Maiti, Dipayan and House, Leanna and Leman, Scotland and North, Chris} }