Thesis (M.S.)--Worcester Polytechnic Institute. / Keywords: anomaly; outlier; visuallization; visual analytics. Includes bibliographical references (leaves 68-72 ).

"Visual analytics is an emerging field which uses visual techniques to interact with users in the analytical reasoning process. Users can choose the most appropriate representation that conveys the important content of their data by acting upon different visual displays. The data itself has many features of interest, including clusters, trends (commonalities) and anomalies. Most visualization techniques currently focus on the discovery of trends and other relations, where uncommon phenomena are treated as outliers and are either removed from the datasets or de-emphasized on the visual displays. Much less work has been done on the visual analysis of outliers, or anomalies. In this thesis, I will introduce a method to identify the different levels of “outlierness” by using interactive selection and other approaches to process outliers after detection. In one approach, the values of these outliers will be estimated from the values of their k-Nearest Neighbors and replaced to increase the consistency of the whole dataset. Other approaches will leave users with the choice of removing the outliers from the graphs or highlighting the unusual patterns on the graphs if points of interest lie in these anomalous regions. I will develop and test these anomaly handling methods within the XMDV Tool."

Approximate query processing (AQP) provides fewer representative samples to approximate large amounts of data. Processing these smaller data subsets enables visualization systems to provide end-users with real-time responses. However, challenges arise for real-world users in adopting AQP-based visualization systems, e.g., the absence of AQP modules in mainstream commercial databases, erroneous estimations caused by sampling bias, and end-user uncertainty when interpreting approximate query results. In this dissertation, we present an AQP-centered technique for enabling interactive visual analytics for large amounts of spatiotemporal data under the aforementioned challenges. First, we design, implement and evaluate a client-based visual analytics framework that progressively acquires spatiotemporal data from an AQP-absence server-side to client-based visualization systems so that interactive data exploration can be maintained on a client machine with modest computational power. Second, we design, implement, and evaluate an online sampling approach that selects samples of large spatiotemporal data in an unbiased manner and accordingly improves the accuracy of the associated estimates. Last, we design, implement and evaluate a difference assessment approach that compares approximate and exact spatial heatmap visualizations in terms of human perception. As such, information changes perceptible by users are well represented, and users can evaluate the reliability of approximate answers more easily. Our results show the superior performance of our proposed AQP-centered technique in terms of speed, accuracy, and user trust, as compared to a baseline of state-of-the-art solutions.

Analysts must filter through an ever-growing amount of data to obtain information relevant to their investigations. Looking at every piece of information individually is in many cases not feasible; there is hence a growing need for new filtering tools and techniques to improve the analyst process with large datasets. We present MineVis — an analytics system that integrates biclustering algorithms and visual analytics tools in one seamless environment. The combination of biclusters and visual data glyphs in a visual analytics spatial environment enables a novel type of filtering. This design allows for rapid exploration and navigation across connected documents. Through a user study we conclude that our system has the potential to help analysts filter data by allowing them to i) form hypotheses before reading documents and subsequently ii) validating them by reading a reduced and focused set of documents. / Master of Science

Sensemaking (i.e. the process of deriving meaning from complex information to make decisions) is often cited as an important and challenging activity for collaborative technology. A key element to the success of collaborative sensemaking is effective coordination and communication within the team. It requires team members to divide the task load, communicate findings and discuss the results. Sensemaking is one of the human activities involved in visual analytics (i.e. the science of analytical reasoning facilitated by interactive visual interfaces). The inherent complexity of the sensemaking process imposes many challenges for designers. Therefore, providing effective tool support for collaborative sensemaking is a multifaceted and complex problem. Such tools should provide support for visualization as well as communication and coordination. Analysts need to organize their findings, hypotheses, and evidence, share that information with their collaborators, and coordinate work activities amongst members of the team. Sharing externalizations (i.e. any information related to the course of analysis such as insights, hypotheses, to-do lists, reminders, etc recorded in the form of note/ annotation) could increase awareness and assist team members to better communicate and coordinate their work activities. However, we currently know very little about how to provide tool support for this sort of sharing. This thesis is structured around three major phases. It consists of a series of studies to better understand collaborative Visual Analytics (VA) processes and challenges, and empirically evaluate design ideas for supporting collaborative sensemaking. I investigate how collaborative sensemaking can be supported during visual analytics by a small team of collocated analysts. In the first phase of this research, I conducted an observational study to better understand the process of sensemaking during collaborative visual analytics as well as identify challenges and further requirements. This study enabled me to develop a deeper understanding of the collocated collaborative visual analytics process and activities involved. I found that record-keeping plays a critical role in the overall process of collaborative visual analytics. Record-keeping involves recording any information related to the analysis task including visualization snapshots, system states, notes, annotations and any other material for further analysis such as reminders and to-do lists. Based on my observations, I proposed a characterization of activities during collaborative visual analytics that encompasses record-keeping as one of the main activities. In addition, I characterized notes according to their content, scope, and usage, and described how they fit into a process of collaborative data analysis. Then, I derived guidelines to improve the design of record-keeping functionality for collocated collaborative visual analytics tools. One of the main design implications of my observational study was to integrate record-keeping functionality into a collaborative visual analytics tool. In order to examine how this feature should be integrated with current VA tools, in the second phase of this research, I designed, developed and evaluated a tool, CoSpaces (Collaborative Spaces), tailor-made for collocated collaborative data analysis on large interactive surfaces. Based on the result of a user study with this tool, I characterized users' actions on visual record-keeping as well as their key intentions for each action. In addition, I proposed further design guidelines such as providing various views of recorded material, showing manually saved rather than automatically saved items by default, enabling people to review collaborators' work unobtrusively, and automatically recommending items related to a user's analytical task. In the third phase, I took supporting record-keeping activities in the context of collaborative sensemaking a step further to investigate how this support should be designed to facilitate collaboration. To this end, I explored how automatic discovery and linking of common work can be employed within a ``collaborative thinking space'' (i.e. a space to enable analysts to record and organize findings, evidence, and hypotheses, also facilitate the process of sharing findings amongst collaborators), to facilitate synchronous collaborative sensemaking activities in visual analytics. The main goal of this phase was to provide an environment for analysts to record, organize, share and connect externalizations. I expected that this would increase awareness among team members and in turn would enhance communication and coordination of activities. I designed, implemented and evaluated a new tool, CLIP (Collaborative Intelligence Pad), that extends earlier thinking spaces by integrating new features that reveal relationships between collaborators' findings. Comparing CLIP versus a baseline tool demonstrated that linking collaborators' work led to significant improvement in analytical outcomes at a collaborative intelligence task. Groups using CLIP were also able to more effectively coordinate their work, and held more discussion of their findings and hypotheses. Based on this study, I proposed design guidelines collaborative VA tools. In summary, I contribute an understanding for how analysts use VA tools during collocated collaboration. Through a series of observational user studies, I investigated how we can better support this complex process. More specifically, I empirically studied recording and sharing of analytical results. For this purpose, I implemented and evaluated two systems to be able to understand the effects of these tools on collaboration mechanics. These user studies along with various literature surveys on each specific topic resulted in a collection of guidelines for supporting and sharing externalizations. In addition, I proposed and evaluated several mechanisms to increase awareness among team members, resulting in more effective coordination and communication during the collaborative sensemaking process. The most novel contributions of this research are the identification and subsequent characterization of note taking behaviours as an important component of visual data exploration and analysis. Moreover, the design and evaluation of CLIP, providing preliminary evidence in support of automatically identifying and presenting relationships between collaborators' findings. / Graduate / 0984 / narges.mahyar@gmail.com

abstract: Causality analysis is the process of identifying cause-effect relationships among variables. This process is challenging because causal relationships cannot be tested solely based on statistical indicators as additional information is always needed to reduce the ambiguity caused by factors beyond those covered by the statistical test. Traditionally, controlled experiments are carried out to identify causal relationships, but recently there is a growing interest in causality analysis with observational data due to the increasing availability of data and tools. This type of analysis will often involve automatic algorithms that extract causal relations from large amounts of data and rely on expert judgment to scrutinize and verify the relations. Over-reliance on these automatic algorithms is dangerous because models trained on observational data are susceptible to bias that can be difficult to spot even with expert oversight. Visualization has proven to be effective at bridging the gap between human experts and statistical models by enabling an interactive exploration and manipulation of the data and models. This thesis develops a visual analytics framework to support the interaction between human experts and automatic models in causality analysis. Three case studies were conducted to demonstrate the application of the visual analytics framework in which feature engineering, insight generation, correlation analysis, and causality inspections were showcased. / Dissertation/Thesis / Doctoral Dissertation Computer Science 2019

The ultimate goal of any visual analytic task is to make sense of the data and gain insights. Unfortunately, the process of discovering useful information is becoming more challenging nowadays due to the growing data scale. Particularly, the human cognitive capabilities remain constant whereas the scale and complexity of data are not. Meanwhile, visual analytics largely relies on human analytic in the loop which imposes challenge to traditional human-driven workflow. It is almost impossible to show every aspect of details to the user while diving into local region of the data to explain phenomenons hidden in the data. For example, while exploring the data subsets, it is always important to determine which partitions of data contain more important information. Also, determining the subset of features is vital before further doing other analysis. Furthermore, modeling on these subsets of data locally can yield great finding but also introduces bias. In this work, a model driven visual analytic framework is proposed to help identify interesting local patterns from the above three aspects. This dissertation work aims to tackle these subproblems in the following three topics: model-driven data exploration, model-driven feature analysis and local model diagnosis. First, the model-driven data exploration focus on the problem of modeling subset of data to identify the co-movement of time-series data within certain subset time partitions, which is an important application in a number of domains such as medical science, finance, business and engineering. Second, the model-driven feature analysis is to discover the important subset of interesting features while analyzing local feature similarities. Within the financial risk dataset collected by domain expert, we discover that the feature correlation among different data partitions (i.e., small and large companies) are very different. Third, local model diagnosis provides a tool to identify interesting local regression models at local regions of the data space which makes it possible for the analysts to model the whole data space with a set of local models while knowing the strength and weakness of them. The three tools provide an integrated solution for identifying interesting patterns within local subsets of data.

When exploring large, high-dimensional datasets, analysts often utilize two techniques for reducing the data to make exploration more tractable. The first technique, dimension reduction, reduces the high-dimensional dataset into a low-dimensional space while preserving high-dimensional structures. The second, clustering, groups similar observations while simultaneously separating dissimilar observations. Existing work presents a number of systems and approaches that utilize these techniques; however, these techniques can cooperate or conflict in unexpected ways. The core contribution of this work is the systematic examination of the design space at the intersection of dimension reduction and clustering when building intelligent, interactive tools in visual analytics. I survey existing techniques for dimension reduction and clustering algorithms in visual analytics tools, and I explore the design space for creating projections and interactions that include dimension reduction and clustering algorithms in the same visual interface. Further, I implement and evaluate three prototype tools that implement specific points within this design space. Finally, I run a cognitive study to understand how analysts perform dimension reduction (spatialization) and clustering (grouping) operations. Contributions of this work include surveys of existing techniques, three interactive tools and usage cases demonstrating their utility, design decisions for implementing future tools, and a presentation of complex human organizational behaviors. / Doctor of Philosophy / When an analyst is exploring a dataset, they seek to gain insight from the data. With data sets growing larger, analysts require techniques to help them reduce the size of the data while still maintaining its meaning. Two commonly-utilized techniques are dimension reduction and clustering. Dimension reduction seeks to eliminate unnecessary features from the data, reducing the number of columns to a smaller number. Clustering seeks to group similar objects together, reducing the number of rows to a smaller number. The contribution of this work is to explore how dimension reduction and clustering are currently being used in interactive visual analytics systems, as well as to explore how they could be used to address challenges faced by analysts in the future. To do so, I survey existing techniques and explore the design space for creating visualizations that incorporate both types of computations. I look at methods by which an analyst could interact with those projections in other to communicate their interests to the system, thereby producing visualizations that better match the needs of the analyst. I develop and evaluate three tools that incorporate both dimension reduction and clustering in separate computational pipelines. Finally, I conduct a cognitive study to better understand how users think about these operations, in order to create guidelines for better systems in the future.

Thesis (Ph. D.)--Washington State University, December 2007. / Includes bibliographical references (p. 98-107).

This thesis examines the usage of and attitudes toward data visualization and analytic tools in three genres of online games. Using an online survey, this research analyzes responses from participants regarding their play habits and attitudes online. Several scales are generated identifying different player demographics such as emotional attitudes, competitive attitudes, technological attitudes, spectator involvement, and overall attitudes toward information customization. In addition, several genre specific scales are created for massive multiplayer online games (MMO), real time strategy (RTS) and first person shooting (FPS) games. This research concludes that competitive attitudes are moderately correlated with information customization and implementation of data visualization tools. Additionally, the relationship between the usage of data visualization tools are strongest with the MMO genre compared to the RTS or FPS genres. In addition, my research shows a strong preference between the responses for the usage of data visualization tools amongst those who report higher levels of spectator involvement with online games. Finally, my research concludes that there is a strong relationship between the amount of time players spend playing online games and the attitudes toward and usage of data visualization tools. / text