{"52390":{"#nid":"52390","#data":{"type":"event","title":"GVU BROWN BAG: InfoVis and Visual Analytics Review Talks","body":[{"value":"\u003Cp\u003EThis Brown Bag reviews papers and posters presented by GT researchers\u003Cbr \/\u003E\nduring the InfoVis Conference and the Visual Analytics Symposium at\u003Cbr \/\u003E\nIEEE VisWeek, October 12-16 in Atlantic City. The following articles\u003Cbr \/\u003E\nwill be presented:\u003C\/p\u003E\n\u003Cp\u003E\u0022SellTrend: Inter-Attribute Visual Analysis of Temporal\u003Cbr \/\u003E\nTransaction Data\u0022 (InfoVis full paper, Honorable Mention Best Paper)\u003Cbr \/\u003E\n\u003Cbr \/\u003E\nAUTHORS:  Zhicheng Liu, John Stasko, Timothy Sullivan\u003C\/p\u003E\n\u003Cp\u003EABSTRACT: We present a case study of our experience designing\u003Cbr \/\u003E\nSellTrend, a visualization system for analyzing airline travel purchase\u003Cbr \/\u003E\nrequests. The relevant transaction data can be characterized as\u003Cbr \/\u003E\nmulti-variate temporal and categorical event sequences, and the chief\u003Cbr \/\u003E\nproblem addressed is how to help company analysts identify complex\u003Cbr \/\u003E\ncombinations of transaction attributes that contribute to failed\u003Cbr \/\u003E\npurchase requests. SellTrend combines a diverse set of techniques\u003Cbr \/\u003E\nranging from time series visualization to faceted browsing and\u003Cbr \/\u003E\nhistorical trend analysis in order to help analysts make sense of the\u003Cbr \/\u003E\ndata. We believe that the combination of views and interaction\u003Cbr \/\u003E\ncapabilities in SellTrend provides an innovative approach to this\u003Cbr \/\u003E\nproblem and to other similar types of multivariate, temporally-driven\u003Cbr \/\u003E\ntransaction data analysis. Initial feedback from company analysts\u003Cbr \/\u003E\nconfirms the utility and benefits of the system.\u003C\/p\u003E\n\u003Cp\u003E\u0022Evaluating Visual Analytics Systems for Investigative Analysis:\u003Cbr \/\u003E\nDeriving Design Principles from a Case Study\u0022 (VAST full paper)\u003Cbr \/\u003E\n\u003Cbr \/\u003E\nAUTHORS:  Youn-ah Kang, Carsten Gorg, John Stasko \u003C\/p\u003E\n\u003Cp\u003EABSTRACT: Despite the growing number of systems providing visual\u003Cbr \/\u003E\nanalytic support for investigative analysis, few empirical studies of\u003Cbr \/\u003E\nthe potential benefits of such systems have been conducted,\u003Cbr \/\u003E\nparticularly controlled, comparative evaluations. Determining how such\u003Cbr \/\u003E\nsystems foster insight and sensemaking is important for their continued\u003Cbr \/\u003E\ngrowth and study, however. Furthermore, studies that identify how\u003Cbr \/\u003E\npeople use such systems and why they benefit (or not) can help inform\u003Cbr \/\u003E\nthe design of new systems in this area. We conducted an evaluation of\u003Cbr \/\u003E\nthe visual analytics system Jigsaw employed in a small investigative\u003Cbr \/\u003E\nsensemaking exercise, and we compared its use to three other more\u003Cbr \/\u003E\ntraditional methods of analysis. Sixteen participants performed a\u003Cbr \/\u003E\nsimulated intelligence analysis task under one of the four conditions.\u003Cbr \/\u003E\nExperimental results suggest that Jigsaw assisted participants to\u003Cbr \/\u003E\nanalyze the data and identify an embedded threat. We describe different\u003Cbr \/\u003E\nanalysis strategies used by study participants and how computational\u003Cbr \/\u003E\nsupport (or the lack thereof) influenced the strategies. We then\u003Cbr \/\u003E\nillustrate several characteristics of the sensemaking process\u003Cbr \/\u003E\nidentified in the study and provide design implications for\u003Cbr \/\u003E\ninvestigative analysis tools based thereon. We conclude with\u003Cbr \/\u003E\nrecommendations for metrics and techniques for evaluating other visual\u003Cbr \/\u003E\nanalytics investigative analysis tools.\u003C\/p\u003E\n\u003Cp\u003E\u0022Two-stage Framework for Visualization of Clustered High Dimensional Data\u0022 (VAST full paper)\u003Cbr \/\u003E\n\u003Cbr \/\u003E\nAUTHORS:  Jaegul Choo, Shawn Bohn, Haesun Park \u003C\/p\u003E\n\u003Cp\u003EABSTRACT: In this paper, we discuss dimension reduction methods for 2D\u003Cbr \/\u003E\nvisualization of high dimensional clustered data. We propose a twostage\u003Cbr \/\u003E\nframework for visualizing such data based on dimension reduction\u003Cbr \/\u003E\nmethods. In the first stage, we obtain the reduced dimensional data by\u003Cbr \/\u003E\napplying a supervised dimension reduction method such as linear\u003Cbr \/\u003E\ndiscriminant analysis which preserves the original cluster structure in\u003Cbr \/\u003E\nterms of its criteria. The resulting optimal reduced dimension depends\u003Cbr \/\u003E\non the optimization criteria and is often larger than 2. In the second\u003Cbr \/\u003E\nstage, the dimension is further reduced to 2 for visualization purposes\u003Cbr \/\u003E\nby another dimension reduction method such as principal component\u003Cbr \/\u003E\nanalysis. The role of the second-stage is to minimize the loss of\u003Cbr \/\u003E\ninformation due to reducing the dimension all the way to 2. Using this\u003Cbr \/\u003E\nframework, we propose several two-stage methods, and present their\u003Cbr \/\u003E\ntheoretical characteristics as well as experimental comparisons on both\u003Cbr \/\u003E\nartificial and real-world text data sets.\u003C\/p\u003E\n\u003Cp\u003E\u0022Social Visualization for Micro-Blogging Analysis\u0022 (InfoVis poster)\u003Cbr \/\u003E\n\u003Cbr \/\u003E\nAUTHORS:  Tanyoung Kim, Hee Young Jeong, Yee Chieh Chew, Matthew Bonner, John Stasko\u003C\/p\u003E\n\u003Cp\u003E\u0022Interactive Visualization of Ecosystem Change for Public Education\u0022 (InfoVis poster)\u003Cbr \/\u003E\n\u003Cbr \/\u003E\nAUTHORS: Tanyoung Kim, Hwajung Hong, Brian Magerko \u003C\/p\u003E\n\u003Cp\u003E\u0022Perspectives on Time: Enhancing Utility with Flexibility\u0022 (InfoVis poster)\u003Cbr \/\u003E\n\u003Cbr \/\u003E\nAUTHORS:  Peter Kinnaird, John Stasko \u003C\/p\u003E\n\u003Cp\u003E\u0022Connect to Congress\u0022 (InfoVis poster)\u003Cbr \/\u003E\n\u003Cbr \/\u003E\nAUTHORS:  by Peter Kinnaird, Hafez Rouzati, Xin Sun\u003C\/p\u003E","summary":null,"format":"limited_html"}],"field_subtitle":"","field_summary":"","field_summary_sentence":"","uid":"27154","created_gmt":"2010-02-11 15:51:36","changed_gmt":"2016-10-08 01:49:39","author":"Louise Russo","boilerplate_text":"","field_publication":"","field_article_url":"","field_event_time":{"event_time_start":"2009-10-29T13:00:00-04:00","event_time_end":"2009-10-29T14:00:00-04:00","event_time_end_last":"2009-10-29T14:00:00-04:00","gmt_time_start":"2009-10-29 17:00:00","gmt_time_end":"2009-10-29 18:00:00","gmt_time_end_last":"2009-10-29 18:00:00","rrule":null,"timezone":"America\/New_York"},"extras":[],"groups":[{"id":"47223","name":"College of Computing"}],"categories":[],"keywords":[],"core_research_areas":[],"news_room_topics":[],"event_categories":[],"invited_audience":[],"affiliations":[],"classification":[],"areas_of_expertise":[],"news_and_recent_appearances":[],"phone":[],"contact":[],"email":[],"slides":[],"orientation":[],"userdata":""}}}