Diane Tang

Polaris: a system for query, analysis, and visualization of multidimensional relational databases

In the last several years, large multidimensional databases have become common in a variety of applications, such as data warehousing and scientific computing. Analysis and exploration tasks place significant demands on the interfaces to these databases. Because of the size of the data sets, dense graphical representations are more effective for exploration than spreadsheets and charts. Furthermore, because of the exploratory nature of the analysis, it must be possible for the analysts to change visualizations rapidly as they pursue a cycle involving first hypothesis and then experimentation. In this paper, we present Polaris, an interface for exploring large multidimensional databases that extends the well-known pivot table interface. The novel features of Polaris include an interface for constructing visual specifications of table-based graphical displays and the ability to generate a precise set of relational queries from the visual specifications. The visual specifications can be rapidly and incrementally developed, giving the analyst visual feedback as he constructs complex queries and visualizations.

Analysis of a local-area wireless network

To understand better how users take advantage of wireless networks, we examine a twelve-week trace of a building-wide local-area wireless network. We analyze the network for overall user behavior (when and how intensively people use the network and how much they move around), overall network traffic and load characteristics (observed throughput and symmetry of incoming and outgoing traffic), and traffic characteristics from a user point of view (observed mix of applications and number of hosts connected to by users). Amongst other results, we find that users are divided into distinct location-based sub-communities, each with its own movement, activity, and usage characteristics. Most users exploit the network for web-surfing, session-oriented activities and chat-oriented activities. The high number of chat-oriented activities shows that many users take advantage of the mobile network for synchronous communication with others. In addition to these user-specific results, we find that peak throughput is usually caused by a single user and application. Also, while incoming traffic dominates outgoing traffic overall, the opposite tends to be true during periods of peak throughput, implying that significant asymmetry in network capacity could be undesirable for our users. While these results are only valid for this local-area wireless network and user community, we believe that similar environments may exhibit similar behavior and trends. We hope that our observations will contribute to a growing understanding of mobile user behavior.

Analysis of a metropolitan-area wireless network

We analyze a seven-week trace of the Metricom metropolitan-area packet radio wireless network to find how users take advantage of a mobile environment. Such understanding is critical for planning future large-scale mobile network infrastructures. Amongst other results, we find that users typically use the radios during the day and evening. Of the users who move around during the trace (over half), we find that the more locations a user visits on a daily basis, the closer together, on average, those locations are. While these results are only known to be valid for this particular network, we hope future analysis of other networks will add to a growing understanding of mobile network behavior.

Overlapping experiment infrastructure: more, better, faster experimentation

At Google, experimentation is practically a mantra; we evaluate almost every change that potentially affects what our users experience. Such changes include not only obvious user-visible changes such as modifications to a user interface, but also more subtle changes such as different machine learning algorithms that might affect ranking or content selection. Our insatiable appetite for experimentation has led us to tackle the problems of how to run more experiments, how to run experiments that produce better decisions, and how to run them faster. In this paper, we describe Googles overlapping experiment infrastructure that is a key component to solving these problems. In addition, because an experiment infrastructure alone is insufficient, we also discuss the associated tools and educational processes required to use it effectively. We conclude by describing trends that show the success of this overall experimental environment. While the paper specifically describes the experiment system and experimental processes we have in place at Google, we believe they can be generalized and applied by any entity interested in using experimentation to improve search engines and other web applications.

Query, analysis, and visualization of hierarchically structured data using Polaris

In the last several years, large OLAP databases have become common in a variety of applications such as corporate data warehouses and scientific computing. To support interactive analysis, many of these databases are augmented with hierarchical structures that provide meaningful levels of abstraction that can be leveraged by both the computer and analyst. This hierarchical structure generates many challenges and opportunities in the design of systems for the query, analysis, and visualization of these databases.In this paper, we present an interactive visual exploration tool that facilitates exploratory analysis of data warehouses with rich hierarchical structure, such as might be stored in data cubes. We base this tool on Polaris, a system for rapidly constructing table-based graphical displays of multidimensional databases. Polaris builds visualizations using an algebraic formalism derived from the interface and interpreted as a set of queries to a database. We extend the user interface, algebraic formalism, and generation of data queries in Polaris to expose and take advantage of hierarchical structure. In the resulting system, analysts can navigate through the hierarchical projections of a database, rapidly and incrementally generating visualizations for each projection.

Multiscale visualization using data cubes

Most analysts start with an overview of the data before gradually refining their view to be more focused and detailed. Multiscale pan-and-zoom systems are effective because they directly support this approach. However, generating abstract overviews of large data sets is difficult and most systems take advantage of only one type of abstraction: visual abstraction. Furthermore, these existing systems limit the analyst to a single zooming path on their data and thus to a single set of abstract views. This paper presents: 1) a formalism for describing multiscale visualizations of data cubes with both data and visual abstraction and 2) a method for independently zooming along one or more dimensions by traversing a zoom graph with nodes at different levels of detail. As an example of how to design multiscale visualizations using our system, we describe four design patterns using our formalism. These design patterns show the effectiveness of multiscale visualization of general relational databases.

Rivet: a flexible environment for computer systems visualization

Rivet is a visualization system for the study of complex computer systems. Since computer systems analysis and visualization is an unpredictable and iterative process, a key design goal of Rivet is to support the rapid development of interactive visualizations capable of visualizing large data sets. In this paper, we present Rivets architecture, focusing on its support for varied data sources, interactivity, composition and user-defined data transformations. We also describe the challenges of implementing this architecture efficiently and flexibly. We conclude with several examples of computer systems visualizations generated within Rivet, including studies of parallel systems, superscalar processors and mobile network usage.

Polaris: a system for query, analysis, and visualization of multidimensional databases

During the last decade, multidimensional databases have become common in the business and scientific worlds. Analysis places significant demands on the interfaces to these databases. It must be possible for analysts to easily and incrementally change both the data and their views of it as they cycle between hypothesis and experimentation. In this paper, we address these demands by presenting the Polaris formalism, a visual query language for precisely describing a wide range of table-based graphical presentations of data. This language compiles into both the queries and drawing commands necessary to generate the visualization, enabling us to design systems that closely integrate analysis and visualization. Using the Polaris formalism, we have built an interactive interface for exploring multidimensional databases that analysts can use to rapidly and incrementally build an expressive range of views of their data as they engage in a cycle of visual analysis.

Experiences with a Mobile Testbed

This paper presents results from an eight-day network packet-trace of MosquitoNet. MosquitoNet allows users of laptop computers to switch seamlessly between a metropolitan-area wireless network and a wired network (10 Mbit/s Ethernet) available in offices and on-campus residences. Results include the amount of user mobility between the wired and wireless networks, the amount of mobility within the wireless network, an examination of application end-to-end delays, and an examination of overall packet loss and reordering in the wireless network. We find that the average mobile host switches between the wired and wireless networks 14 times during the trace and moves within the wireless network five times. Round trip latencies in the wireless network are very high, with a minimum of 0.2 seconds. Even higher end-to-end delays, of up to hundreds of seconds, are due to packet loss and reordering. These delays cause users to change their usage patterns when connected to the wireless network. We conclude that latency is a critical problem in the wireless network.

Design choices when architecting visualizations

In this paper, we focus on some of the key design decisions we faced during the process of architecting a visualization system and present some possible choices, with their associated advantages and disadvantages. We frame this discussion within the context of Rivet, our general visualization environment designed for rapidly prototyping interactive, exploratory visualization tools for analysis. As we designed increasingly sophisticated visualizations, we needed to refine Rivet in order to be able to create these richer displays for larger and more complex data sets. The design decisions we discuss in this paper include: the internal data model, data access, semantic meta-data information the visualization can use to create effective visual decodings, the need for data transformations in a visualization tool, modular objects for flexibility, and the tradeoff between simplicity and expressiveness when providing methods for creating visualizations.