Computer graphics processing and selective visual display system – Display driving control circuitry – Controlling the condition of display elements
Reexamination Certificate
2000-05-16
2002-04-02
Sax, Steven (Department: 2173)
Computer graphics processing and selective visual display system
Display driving control circuitry
Controlling the condition of display elements
C345S678000, C345S663000, C345S586000
Reexamination Certificate
active
06366299
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention is directed to a method of information interaction that allows users of computers to more effectively browse multidimensional information and understand hidden relationships contained in that information.
2. Related Art
The basic prior art paradigms for user interaction with and finding of information have not changed significantly in the commercial marketplace for years. In a so-called standard browsing paradigm, users follow hyperlinks, i.e., links to other documents embedded within Hypertext Markup Language (HTML) documents, the type of documents found, for example, on the World Wide Web (the “Web”). The paradigm is named after the popular term “browsing the Web” for navigating the Web using browser software, such as Netscape Navigator or Microsoft Explorer. Such browsing typically begins with broad categories and terminates with detailed listings. False starts require users to back up and try again with often confusing navigation mechanisms.
Another such paradigm will be referred to as a query/response paradigm, used for example when searching using an Internet search engine or index, in which users enter a query and then the system responds with an ordered list of results. Often users have to repeat this pattern several times to refine the query.
Examples of each of the above-mentioned interaction paradigms are ubiquitous, particularly on the Internet.
FIG. 1
shows an extended directory example of the query/response paradigm from an online service known as PeopleLink (available at the time of filing at: www.peoplelink.com). The intention of PeopleLink is to support users in finding chat partners who may share their interests. The interaction paradigm here is much the same as any other standard query/response problem.
A user begins by specifying a query by choosing from selection tools, or “widgets” that include drop-down lists and check boxes. Results from the search include a listing of potential partners at reference numeral
2
, ordered in this case alphabetically. The listings themselves include links to related information and relevant attribute values, such as whether the individual is online at the time, at reference numeral
3
as well as links to profiles at reference numeral
4
of the listed people.
In this particular example, 3458 matches for the query were found, of which the first 20 are shown at reference numeral
5
. There are usually two strategies available for users to proceed from this point: (1) back up and refine the search and/or (2) evaluate the candidate items one by one by flipping through listings pages.
Page flipping and scanning through long lists of result items is tedious and time-consuming, as is query refinement. The problem is magnified when the user is not sure what he or she is looking for. Note that in the directory lookup problem to which the technique of
FIG. 1
is directed it is not a trivial matter to determine whether the goal has been reached. It is not at all clear how a user would know if he or she has found a satisfactory chat partner. That determination may depend on what the total set of choices is and how they compare across the attributes available. This type of problem is known as a semi-structured problem, as opposed to a straightforward search of, say, phone numbers by name. Such problems are common in the field of information access, particularly in the shopping domain, and also in many others.
Various information visualization approaches have previously been brought to bear on the information access problem. Such approaches have to date come largely from within academic research communities. One class of prior art solutions to improving on the basic query/response technique has been to seek methods to anticipate search results before issuing the full query, thus speeding the query refinement loop, and improving a query's probability of success. A first step in this technique is to indicate the size of the item set before users issue a query, thus giving the user a hint about how much he or she needs to restrict the set beyond its initial size before going into an evaluation/browsing mode. An even better technique is to provide instant feedback after each part of a query is formed. Such techniques are the basis of Dynamic Querying, a paradigm developed at the University of Maryland. A description of Dynamic Querying can be found in Shneiderman, B., “Dynamic Queries, Starfield Displays, and the Path to Spotfire” (1999).
The basic Dynamic Querying technique uses sliders and other widgets to formulate queries; visual feedback is available instantly at a meta-level on the results of the queries in a separate visualization. Early prototypes used a scatterplot to present abstract query results, but later commercializations have used a variety of visualization techniques.
FIG. 2
shows a screendump from a commercial product called Spotfire, from a company founded by one of the students from the University of Maryland, that utilizes the dynamic querying technique. Queries are made with the sliders
10
at right. Results are visualized with the two-dimensional plots and graphs
12
on the left. Based on information available at the company's website, this product seems to have found its niche in scientific applications. Its sophisticated user interface is not designed to be used by the casual user for, say, shopping, although many earlier prototypes built at the University of Maryland were intended to address more casual users. Shneiderman comments, in the article cited above, on the puzzling fact that these interface ideas have not found easy acceptance within commercial applications such as real estate and movie databases.
It does not appear in any event that these commercial sectors have acknowledged the significance of the interface in the information access problem. One possible factor is that the implementations typically have not been built with mass-market Web browsers in mind. In fact, many of the basic performance goals would be difficult to achieve on the Web without specialized browser plugins.
A group headquartered at the Imperial College of London integrates the query mechanism with the results display using dynamic histograms. The research prototypes Attribute Explorer and Influence Explorer use interactive histograms to explore the relationships of attributes in a given set. These techniques are described in several articles: Tweedie, L. A., Spence, R., Dawkes, H., and Su, H. (1996) “Externalizing Abstract Mathematical Models”, Proceedings of CHI ′96, Vancouver, Canada, ACM Press, pp. 406-412; and Tweedie, L. A., R. Spence, D. Williams, and R. Bhogal (1994) “The Attribute Explorer”, Video Proceedings and Conference Companion of CHI ′94, New York, ACM Press, pp. 435-436.
FIG. 3
shows how Attribute Explorer makes use of synchronized interactive histograms to make queries and exhibit attribute relationships in a data set. This example domain relates to housing. In the example shown in the figure, there are three attributes: price
20
, distance from a certain location
21
, and size of the garden (lawn)
22
. Histograms are used in all cases to show the distribution of the data across these attributes. A user can restrict an attribute's value range by interacting with the sliders at the bottom of the windows; results are painted back onto all the displays by coloring the histograms.
Interactive histograms have the advantage of showing aggregate relationships clearly. The display is by value rather than by item. That is, a value range is plotted on an axis and the number of items that fall into that value is plotted on the second axis to yield a histogram. This is done for each attribute dimension.
In the area of evaluation of result items, many existing sites have turned to methods for direct attribute comparison through tables. The more advanced of these may allow users to sort the tables interactively by different attributes. A good example of a shopping site that has some of these table comparison features
Fyock Christina Lynn
Heinrichs Michael Scott
Lanning Thomas Richard
Li Guangjun
Wittenburg Kent Barrows
Sax Steven
Suchyta Leonard Charles
Verizon Laboratories Inc.
Weixel James K.
LandOfFree
Multidimensional information visualization using attribute rods does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Multidimensional information visualization using attribute rods, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Multidimensional information visualization using attribute rods will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2819722