STREAMIT: Dynamic visualization and interactive exploration of text streams

Abstract

Text streams demand an effective, interactive, and on-the-fly method to explore the dynamic and massive data sets, and meanwhile extract valuable information for visual analysis. In this paper, we propose such an interactive visualization system that enables users to explore streaming-in text documents without prior knowledge of the data. The system can constantly incorporate incoming documents from a continuous source into existing visualization context, which is "physically" achieved by minimizing a potential energy defined from similarities between documents. Unlike most existing methods, our system uses dynamic keyword vectors to incorporate newly-introduced keywords from data streams. Furthermore, we propose a special keyword importance that makes it possible for users to adjust the similarity on-the-fly, and hence achieve their preferred visual effects in accordance to varying interests, which also helps to identify hot spots and outliers. We optimize the system performance through a similarity grid and with parallel implementation on graphics hardware (GPU), which achieves instantaneous animated visualization even for a very large data collection. Moreover, our system implements a powerful user interface enabling various user interactions for in-depth data analysis. Experiments and case studies are presented to illustrate our dynamic system for text stream exploration.

Semantic Interaction for Visual Analytics: Inferring Analytical Reasoning for Model Steering

“…Alsakran et al presented a visualization system, STREAMIT, capable of spatially arranging text streams based on keyword similarity [3]. Again, users can interactively explore and adjust the spatial layout through directly changing the weight of keywords that they find important.…”

Semantic Interaction for Sensemaking: Inferring Analytical Reasoning for Model Steering

“…Users can interact and modify this view via direct manipulation of the keyword weighting (i.e., if a user finds a term more important, she can increase the weight on that term). Similarly, STREAMIT enables direct control over keyword weighting in a force-directed model [10]. Visualizations such as "Dust & Magnet" [11] and VIBE [12] allow users to place specific points of interest in the space, acting as anchors from which the remaining points attract and repel form.…”

Supporting effective common ground construction in Asynchronous Collaborative Visual Analytics

“…Note that the two insights can be about different datasets that share dimensions and data items. By considering each dimension as a weighted keyword and an insight as a document, we use an improved cosine similarity measure [3] to calculate Sim dim :…”

“…The insights are clustered according to their correlations. Please refer to [3] for details of the underlying force-based dynamic system of this view.…”

“…For example, by setting the tag weight to 1 through the star glyph (see Figure1(4)), insights are grouped by their tags. Users can also interactively adjust the importance of keywords from the keyword table 3 (see Figure 1(5)) to cluster the insights by keywords of interest, 3 Keywords stored in this table are insight contents, such as dimension as shown in Figure 1(1). Animation: Users can play animations to examine temporal evolution of the insights.…”