Real-world data is known to be imperfect, suffering from various forms of defects such as sensor variability, estimation errors, uncertainty, human errors in data entry, and gaps in data gathering. Analysis conducted on variable quality data can lead to inaccurate or incorrect results. An effective visualization system must make users aware of the quality of their data by explicitly conveying not only the actual data content, but also its quality attributes. While some research has been conducted on visualizing uncertainty in spatiotemporal data and univariate data, little work has been reported on extending this capability into multivariate data visualization. In this paper we describe our approach to the problem of visually exploring multivariate data with variable quality. As a foundation, we propose a general approach to defining quality measures for tabular data, in which data may experience quality problems at three granularities: individual data values, complete records, and specific dimensions. We then present two approaches to visual mapping of quality information into display space. In particular, one solution embeds the quality measures as explicit values into the original dataset by regarding value quality and record quality as new data dimensions. The other solution is to superimpose the quality information within the data visualizations using additional visual variables. We also report on user studies conducted to assess alternate mappings of quality attributes to visual variables for the second method. In addition, we describe case studies that expose some of the advantages and disadvantages of these two approaches.

Visualization systems traditionally focus on graphical representation of information. They tend not to provide integrated analytical services that could aid users in tackling complex knowledge discovery tasks. Users' exploration in such environments is usually impeded due to several problems: 1) valuable information is hard to discover when too much data is visualized on the screen; 2) Users have to manage and organize their discoveries off line, because no systematic discovery management mechanism exists; 3) their discoveries based on visual exploration alone may lack accuracy; and 4)they have no convenient access to the important knowledge learned by other users. To tackle these problems, it has been recognized that analytical tools must be introduced into visualization systems. In this paper, we present a novel analysis-guided exploration system, called the Nugget Management System (NMS). It leverages the collaborative effort of human comprehensibility and machine computations to facilitate users' visual exploration processes. Specifically, NMS first helps users extract the valuable information (nuggets) hidden in datasets based on their interests. Given that similar nuggets may be rediscovered by different users, NMS consolidates the nugget candidate set by clustering based on their semantic similarity. To solve the problem of inaccurate discoveries, localized data mining techniques are applied to refine the nuggets to best represent the captured patterns in datasets. Visualization techniques are then employed to present our collected nugget pool and thus create the nugget view. Based on the nugget view, interaction techniques are designed to help users observe and organize the nuggets in a more intuitive manner and eventually faciliate their sense-making process. We integrated NMS into XmdvTool, a freeware multivariate visualization system. User studies were performed to compare the users' efficiency and accuracy in finishing tasks on real datasets, with and without the help of NMS. Our user studies confirmed the effectiveness of NMS.

Abstract. When using visualization techniques to explore data streams, an important task is to convey pattern changes. Challenges include: (1) Most data analysis tasks require users to observe the pattern change over a long time range; (2) The change rate of patterns is not a constant, and most users are normally more interested in bigger changes than smaller ones. Although distorting the time axis as proposed in the literature can partially solve this problem, most of these are driven by the user. This is however not applicable to streaming data exploration tasks that normally require near real-time responsiveness. In this paper, we propose a data-driven framework to merge and thus condense time windows having small or no changes. Only significant changes are shown to users. Juxtaposed views are discussed for conveying data pattern changes. Our experiments show that our merge algorithm preserves more change information than uniform sampling. We also conducted a user study to confirm that our proposed techniques can help users find pattern changes more quickly than via a non-distorted time axis.

We will demonstrate our system, called V iStream, supporting interactive visual exploration of neighbor-based patterns [7] in data streams. V istream does not only apply innovative multi-query strategies to compute a broad range of popular patterns, such as clusters and outliers, in a highly efficient manner, but it also provides a rich set of visual interfaces and interactions to enable real-time pattern exploration. In our demonstration, we will illustrate that with ViStream, analysts can easily interact with the pattern mining processes by navigating along the time horizons, abstraction levels and parameter spaces, and thus better understand the phenomena of interest.

Visual analytics, Quality measures, Information loss,