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.
When data analysts study time-series data, an important task is to discover how data patterns change over time. If the dataset is very large, this task becomes challenging. Researchers have developed many visualization techniques to help address this problem. However, little work has been done regarding the changes of multivariate patterns, such as linear trends and clusters, on time-series data. In this paper, we describe a set of history views to fill this gap. This technique works under two modes: merge and non-merge. For the merge mode, merge algorithms were applied to selected time windows to generate a change-based hierarchy. Contiguous time windows having similar patterns are merged first. Users can choose different levels of merging with the tradeoff between more details in the data and less visual clutter in the visualizations. In the non-merge mode, the framework can use natural hierarchical time units or one defined by domain experts to represent timelines. This can help users navigate across long time periods. Gridbased views were designed to provide a compact overview for the history data. In addition, MDS pattern starfields and distance maps were developed to enable users to quickly investigate the degree of pattern similarity among different time periods. The usability evaluation demonstrated that most participants could understand the concepts of the history views correctly and finished assigned tasks with a high accuracy and relatively fast response time.
Design patterns have proven to be a useful means to make the process of designing, developing, and reusing software
systems more efficient. In the area of information visualization, researchers have proposed design patterns for different
functional components of the visualization pipeline. Since many visualization techniques need to display derived data as
well as raw data, the data transformation stage is very important in the pipeline, yet existing design patterns are, in general,
not sufficient to implement these data transformation techniques. In this paper, we propose two design patterns, operatorcentric
transformation and data modifier, to facilitate the design of data transformations for information visualization
systems. The key idea is to use operators to describe the data derivation and introduce data modifiers to represent the
derived data. We also show that many interaction techniques can be regarded as operators as defined here, thus these two
design patterns could support a wide range of visualization techniques. In addition, we describe a third design pattern,
modifier-based visual mapping, that can generate visual abstraction via linking data modifiers to visual attributes. We also
present a framework based on these three design patterns that supports coordinated multiple views. Several examples of
multivariate visualizations are discussed to show that our design patterns and framework can improve the reusability and
extensibility of information visualization systems. Finally, we explain how we have ported an existing visualization tool
(XmdvTool) from its old data-centric structure to a new structure based on the above design patterns and framework.
The scatterplot matrix is one of the most common methods used to project multivariate data onto two dimensions for display. While each off-diagonal plot maps a pair of non-identical dimensions, there is no prescribed mapping for the diagonal plots. In this paper, histograms, 1D plots and 2D plots are drawn in the diagonal plots of the scatterplots matrix. In 1D plots, the data are assumed to have order, and they are projected in this order. In 2D plots, the data are assumed to have spatial information, and they are projected onto locations based on these spatial attributes using color to represent the dimension value. The plots and the scatterplots are linked together by brushing. Brushing on these alternate visualizations will affect the selected data in the regular scatterplots, and vice versa. Users can also navigate to other visualizations, such as parallel coordinates and glyphs, which are also linked with the scatterplot matrix by brushing. Ordering and spatial attributes can also be used as methods of indexing and organizing data. Users can select an ordering span or a spatial region by interacting with 1D plots or with 2D plots, and then observe the characteristics of the selected data subset. 1D plots and 2D plots provide the ability to explore the ordering and spatial attributes, while other views are for viewing the abstract data. In a sense, we are linking what are traditionally seen as scientific visualization methods with methods from the information visualization and statistical graphics fields. We validate the usefulness of this integration by providing two case studies, time series data analysis and spatial data analysis.
This paper presents NVIS, an interactive graphical tool used to examine the weights, topology, and activations of a single artificial neural networks (ANN), as well as the genealogical relationships between members of a population of ANNs as they evolve under an evolutionary algorithm. NVIS is unique in its depiction of nodal activation values, its usage of family tree diagrams to indicate the origin of individual networks, and the degree of interactivity it allows the user while the learning process takes place. The authors have made use of these feature to obtain insights into both the workings of single neural networks and the evolutionary process, based upon which we consider NVIS to be an effective visualization tool of value to designers, users, and students of ANNs.
Image segmentation systems which employs domain specific knowledge produce, in general, results superior to context-free systems. A significant drawback with using domain specific knowledge is its lack of portability and excessive development and computation time. In this paper, we investigate using a general rule based segmentation system and augmenting it with domain constraint knowledge to improve the performance of the system. We also investigate interactions between multiple types of domain constraint knowledge and their effect on parameter selection and rule ordering.
Image segmentation techniques which employ domain knowledge produce in general much better results than context-free methods but often suffer from lack of portability and excessive development and computation time. We describe a new segmentation technique which employs very general knowledge of domain characteristics to improve the performance of context-free systems while avoiding many of the problems of existing domain-specific methods.