CATEGORY: Data Mining & Visualization

REVIEWER: Jim Acker, senior system analyst for Frank Russell Company.

BACKGROUND: Frank Russell Company, a global investment services firm, provides manager-of-manager investment products and services in more than 35 countries. Russell manages $66 billion in assets and advises clients representing more than $1 trillion (U.S.) worldwide. Founded in 1936, Russell is a subsidiary of Northwestern Mutual and is headquartered in Tacoma, Washington.

PLATFORM: Sun 4500.

PROBLEM SOLVED: When looking at large amounts of data, graphical displays provide a perspective of the data set that can't be achieved with tabular displays. At the same time, tabular displays provide a perspective of individual data points that can't be achieved with graphical displays. Inxight's Table Lens provides a graphical display that offers both perspectives in a single view. The user can see the overall data set, while also singling out individual data points. Table Lens gives users a powerful way to visually identify outliers and patterns in their data.

PRODUCT FUNCTIONALITY: At Russell, we use the Table Lens software development kit (SDK) to connect the applet to various data sources, including Sybase, FAME and SAS. The data connection is made by a JSP page, which calls a Java bean that knows how to convert the data into XML data that the applet can consume. To date, we have largely used the applet as a standalone tool for looking at specific data sets. Moving forward, we plan to use it as a common display option for reports and other applications.

STRENGTHS: The most useful feature of the product is its ability to display an entire data set on one screen. This allows users to easily identify outliers, patterns and trends in their data. Additional features include the ability to drag and drop columns into groups and to identify quartiles. The grouping feature is very useful for sub-setting our data for easier visual analysis. We frequently look at information in quartile space, and Table Lens has the ability to divide the data into quartiles using the menu-bar tools.

WEAKNESSES: The evolution of the product has focused on the applet and left the problem of dynamic data display to the programmer and the SDK. It would be helpful if Inxight enhanced the API to help programmers develop dynamic data feeds to the applet. We are using an XML format used in a demonstration applet that we received from Inxight, which introduces tagging overhead in larger data sets. We need to develop ways to work around this so we can display larger data sets in Table Lens.

SELECTION CRITERIA: We selected Inxight's Table Lens in addition to other visualization tools for use on Russell's Web site. The product's ability to present all the data to the user on one page is unique in the marketplace. We also like that our users immediately understand how the tool can be applied to their requirements. It is important to us that the applet is programmable and provides a well-documented API. The Table Lens SDK meets these requirements and allows us to configure the applet to our specific data requirements.

DELIVERABLES: At Russell, we use Table Lens to look at portfolio data. We load portfolio characteristics or stock characteristics for many portfolios at the same time. Our analysts look at this information to identify exceptions and patterns across portfolios, based on these characteristics. We are looking at using the grouping feature to group time-series data by time period. By using this feature for time and portfolio, we can view portfolio characteristics for specific portfolios, or groups of portfolios, across time.

VENDOR SUPPORT: The product is very easy to install; therefore, support is seldom required. We have contacted Inxight with questions about the product and the SDK. They have always been responsive to our inquiries without delay and to our satisfaction.

DOCUMENTATION: The principle documentation we use is the SDK, which includes good documentation for developing Table Lens applications. The end-user help files are also good references and come in the form of a prebuilt set of customizable Web pages. Some of the documentation is available in HTML and PDF formats.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access