Models and methods for reducing visual complexity and search effort via ideal information abstraction, hiding, and sequencing

• US 20040267600A1
• Filed: 06/30/2003
• Published: 12/30/2004
• Est. Priority Date: 06/30/2003
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Horvitz, Eric J.
• Granted Patent: US 8,346,587 B2
• Time in Patent Office: Days
• US Class Current: 705/7.31
• CPC Class Codes: G06F 16/954 Navigation, e.g. using categorised browsing
• CPC Class Codes: G06Q 30/0202 Market predictions or forecasting for commercial activities
• CPC Class Codes: G06Q 30/0204 Market segmentation

Abstract

The present invention relates to a system and methodology for controlling and presenting information to users in an automated manner. Sets of information items accessible by users from computer workspaces (e.g., from within and/or outside an application) are automatically hidden or reduced from view in order to facilitate user processing of a determined and/or reduced subset of information. The reduced subset is created from decision-theoretic considerations of the expected costs and benefits of hiding or mitigating information items from view or presentation while leaving more likely items or objects exposed for efficient access by users. The display of the ideal set of items takes into consideration controls that allow for the access of views that reveal sets of less likely items through one or more gestures. The control of the optimization may be facilitated by preference-assessment interfaces that allow a designer and/or a user to assess preferences about the costs of reviewing sets of items of different lengths and about the costs of gestures to reveal additional items.

Claims

1. A system that facilitates displaying information, comprising:

a ranking component that determines a likelihood value for at least one display item;

a cost component that determines a navigation cost for the display item; and

a hiding component that selectively exposes a subset of display items from a plurality of display items based at least in part upon the determined likelihood value and the navigation cost.

2. The system of claim 1, the hiding component includes an expected utility analysis component to segment a set of candidate items a into a subset of displayed items d and one or more subsets of items h that are hidden but accessible via action or gesture, wherein a, d, and h are integers respectively.

3. The system of claim 2, the action or gesture includes a mouse click or keystroke to navigate for hidden information.

4. The system of claim 2, the hiding component processes the set a that includes a number of elements |a|=n total candidate items that are potentially of interest or value to a user, the n items are automatically separated into proper subsets of a, including a displayed subset d of |d| items and a hidden subset h of |h| items, where |d|+|h|=n.

5. The system of claim 1, the ranking component includes a function that applies an ordering to the likelihood of items to a user in a context, defined by observed evidence, E.

6. The system of claim 5, the observed evidence E includes one or more of background information about usage, usage patterns of items or related items, content associated with items, time of day, content topic, recently accessed content or recently and accessed topic content.

7. The system of claim 5, the ranking component includes a ranking function as a coarse or accurate estimate of a likelihood that each item i in a set a is a target item at a time of display and/or access, denoted as p(i is target|E), wherein p is a probability.

8. The system of claim 7, the ranking component includes a probability function that maps to other ranking functions that are learned or applied as a heuristic function.

9. The system of claim 8, the probability function determines a likelihood, p(i is target t|E), the probability function is normalized to 1.0 for all of the items stated as:

10. The system of claim 2, the hiding component determines for any set of items selected to be moved to a set of displayed items, d, a likelihood that a target item t will be in the set is stated as:

11. The system of claim 2, the hiding component determines an expected value including expected costs and benefits of moving the top items of d, leading to a cardinality of |d|, and the remaining |h| into the set of hidden items, h, where the total items in an original set a, |a|=|d|+|h|.

12. The system of claim 2, the hiding component determines at least one cost as follows:

C_s,d;

Cost of a user scanning through d given target is in d;

C_s,h;

Cost of a user scanning through h (when set of items h is accessed);

C_h;

Cost of a user accessing h;

C_s,d′

;

Cost of a user scanning through d given that d does not contain the target; and

C₀;

Cost of obfuscation of other content as a function of list, or other configuration of the items, size.

13. The system of claim 12, the costs can be summarized by processing functions that yield estimates of the costs for different situations based on details as specifics of a configuration.

14. The system of claim 13, further comprising functions developed through observation and fitting of time and/or frustration of users as gleaned from user studies, or via heuristic estimates of the time, providing a set of functions that assign cost.

15. The system of claim 12, the cost, C_s,d is associated with scanning items in d, which is a function of configuration and g(|d|).

16. The system of claim 15, further comprising a list configuration having an expected cost (on average) of finding a target item on the list, when the target item is present, the expected cost is measured in terms of the number of items that are scanned before finding the target, is |d|/2 if the list is random, and if the list is sorted by likelihood (versus randomly), employ |d|/2 as a average-case expectation.

17. The system of claim 15, the cost is an average cost of scanning down a list to find a target item, given a list sorted from a highest to lowest probability that each item is the target item and described by the following equations:

p(x_i|E) as represents a likelihood p(item x_i is the target t|E). The expected number of items <

n>

to scan before finding target,

$<<br><br>n><br><br>=p<br><br>\left(x_1<br><br>E\right)*1+\left(1-p<br><br>\left(x_1<br><br>E\right)\right)*p<br><br>\left(x_2<br><br>E\right)*2+\left(1-p<br><br>\left(x_1<br><br>E\right)\right)*\left(1-p<br><br>\left(x_2<br><br>E\right)\right)*p<br><br>\left(x_3<br><br>E\right)*3+\left(1-p<br><br>\left(x_1<br><br>E\right)\right)*\left(1-p<br><br>\left(x_2<br><br>E\right)\right)<br><br>\left(1-p<br><br>\left(x_3<br><br>E\right)\right)*p<br><br>\left(x_4<br><br>E\right)*4<br><br>\mathrm{\dots <br><br>}+\left(1-p<br><br>\left(x_1<br><br>E\right)\right)*\left(1-p<br><br>\left(x_2<br><br>E\right)\right)<br><br>(1-p<br><br>\left(x_3<br><br>E\right)<br><br>\text{\hspace{1em}<br><br>}<br><br>\mathrm{\dots <br><br>}<br><br>\text{\hspace{1em}<br><br>}<br><br>\left(1-p<br><br>\left(x_n<br><br>E\right)\right)*p<br><br>\left(x_n<br><br>E\right)*n.$

18. The system of claim 17, the cost is stated as the following equation:

19. The system of claim 18, the equation is weighted by considering a time associated with scanning items, substituting t(n), for a cost of scanning n items as follows,

20. The system of claim 12, further comprising a function capturing the cost of scanning increasingly longer lists, C(n), is provided as a nonlinear function that grows exponentially with length as follows:

21. The system of claim 12, the cost includes a cost of performing a gesture to make access to hidden items occur, C_h, which is a function of a user interface component.

22. The system of claim 21, the user interface component exposes h after a pre-set threshold of dwell time on d including a cost that is a function of the time associated with a dwell for a threshold time greater than the pre-set threshold.

23. The system of claim 22, the cost includes a time and/or frustration with performing an explicit gesture.

24. The system of claim 21, the user interface component processes a model of cost to the user where the cost of scanning items grows linearly with the number of items, the user interface component sets a cost of access equivalent to scanning a predetermined number of items.

25. The system of claim 12, the cost is computed as an overall cost associated with a specific segmentation as follows:

26. The system of claim 25, further comprising an ideal segmentation computation that is determined by identifying an ideal d, d, of size |d|*, the segmentation of all items a into the displayed items d (and h via complementarity) that minimizes a cost to the user as follows:

27. The system of claim 1, the display item resides on a workspace including at least one of inside an application and outside the application.

28. The system of claim 1, the display item includes at least one of icons, folders, directories, lists, menus, graphics, text, data, and applications.

29. A computer readable medium having computer readable instructions stored thereon for implementing at least one of the ranking component and the cost component of claim 1.

30. A method that facilitates displaying information, comprising:

identifying information that can be displayed on a computer; and

selectively displaying a subset of information from a set of information based at least in part upon an expected utility analysis.

32. The method of claim 30, further comprising automatically determining navigation costs to relocate or transform a display item in view of the assigned rank.

33. The method of claim 30, further comprising automatically determining whether a cost of navigating to a hidden item is outweighed by a benefit of actually hiding the item.

34. The method of claim 30, further comprising performing automated item hiding operations during at least one of a background task, a foreground task, an application task, and an operating system task.

31. The method of clam 30, further comprising automatically assigning a rank to one or more display items based upon one or more relevance variables.

35. A system that facilitates item presentation, comprising:

means for automatically determining a cost to display an item; and

means for selectively exposing the item based upon the determined cost to display the item.

36. A user interface for selectively exposing information to a user, comprising;

a component to receive user feedback related to a cost to access a document and a cost to access non-displayed documents; and

a component to display objects to a user based at least in part on the user feedback.

37. The user interface of claim 36, further comprising a search engine having non-displayed objects viewable by a selection option.

38. The user interface of claim 36, further comprising a component that employs decomposition of a into d and h that are applied recursively to enable multiple level hiding for a nesting of analyses.