Flexible rule-based recommender system for use with an electronic board system

• US 20050021518A1
• Filed: 08/04/2004
• Published: 01/27/2005
• Est. Priority Date: 12/22/2000
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Inventors: Grasso, Antonietta, Snowdon, Dave, Bentley, Richard, Pycock, James
• Granted Patent: US 7,287,024 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 1/16 Constructional details or arrangements
• CPC Class Codes: G06F 3/042 by opto-electronic means
• CPC Class Codes: G06F 3/0481 based on specific properties of the displayed interaction object or a metaphor-based environment, e.g. interaction with desktop elements like windows or icons, or assisted by a cursor's changing behaviour or appearance
• CPC Class Codes: G06F 3/0488 using a touch-screen or digitiser, e.g. input of commands through traced gestures
• CPC Class Codes: G06Q 10/10 Office automation; Time management
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching
• CPC Class Codes: Y10S 715/971 Cooperative decision support systems for group of users

Abstract

An electronic board system includes an electronic board having a screen for displaying information of interest to a work group or community, an input device for receiving information from users in a group or community to be displayed on the electronic board, a memory for storing information received from the input device and a processor for selecting which information stored in the memory to display on the screen, where and how to display the selected information on the screen and displaying the selected information on the screen. The processor selects which information to display in accordance with a predetermined relationship based on a group-based recommendation criteria and user interest. A plurality of sensors disposed behind the screen detect user interest in information displayed on the screen near the sensor. User interest may also be determined by monitoring user requests for copies of information displayed on the screen.

Claims

1. A method for selecting items to display on an electronic board system, wherein the electronic board system includes a screen for displaying items, comprising:

for each user in a community of users having a set of contexts to which the users subscribe, determining the number of interactions with items in contexts the user is a member and the number of interactions with items in contexts the user is not a member;

for each context, evaluating the number and type of user interactions according to a first predetermined relationship;

determining the most popular contexts;

for each of the most popular contexts, selecting the most popular items according to a second predetermined relationship for evaluating user interactions with items; and

displaying the most popular items in each context on the screen.

2. The method of claim 1, wherein the user interactions with items comprise input interactions and output interactions.

3. The method of claim 2, wherein an input interaction includes one of posting an item, commenting on an item, and reviewing an item;

and wherein an output interaction includes one of reading an item, printing an item, copying an item and downloading an item.

4. The method of claim 2, further comprising:

constructing a vector representing a user'"'"'s overall activity of a user,

A_i^j|_j=0<sup>n−

1</sup>=|OF_i<sup>jn−

1</sup>_j=0|+|IF_i^j|_j=0<sup>n−

1</sup>

where

${<br><br>{\mathrm{IF}}_i^j<br><br>}_{j=0}^{n-1}=\underset{l=0}{\overset{m-1}{\sum <br><br>}}<br><br>{\mathrm{dr}}_{\mathrm{jl}}$

is a matrix of contexts by items with input interactions, where each entry dr_ij is 1 if the user introduced the item j in the community i and 0 otherwise, where 0<

j<

n, n=total number of contexts and where 0<

l<

m, m=total number of items submitted to the system,

and

${<br><br>{\mathrm{OF}}_i^j<br><br>}_{j=0}^{n-1}=\underset{l=0}{\overset{m-1}{\sum <br><br>}}<br><br>{\mathrm{dv}}_{\mathrm{jl}}$

is a matrix of contexts by items with output interactions, where each entry dv_ij is 1 if the user followed the item j in the context of the community i and 0 otherwise, where 0<

j<

n, n=total number of contexts and where 0<

1<

m, m=total number of items submitted in the system.

5. The method of claim 1, further comprising:

determining the items having the most recent user interaction; and

for each of the most popular contexts, selecting the most relevant items according to a third predetermined relationship for evaluating the most popular items and the most recent items; and

displaying the most relevant items in each context on the screen.

6. The method of claim 1, further comprising applying a different second predetermined relationship for each item type.

7. The method of claim 2, further comprising a weight of W_i to each input interaction and a weight of W_o to each output interaction.

8. The method of claim 1, wherein the size and location of items displayed on the screen is a function of time displayed and user interest.

9. A method for selecting items to display on an electronic board system, wherein the electronic board system includes a screen for displaying items, comprising:

providing items and recommendations for items stored in the electronic board system from users in a community having at least one context to which the users subscribe;

monitoring user activity with the electronic board system, wherein user activity comprises the number of times individual users have input an item to the electronic board system and the number of times individual users have output the item from the system;

determining which items to be displayed on the screen according to a backstage rule, wherein the backstage rule prioritizes items based on most popular context and most user activity; and

determining the size and location of the selected items on the screen in accordance with a frontstage rule.

10. The method of claim 9, wherein the frontstage rule is a function of time displayed and user interest.

11. The method of claim 9, wherein the backstage rule is a function of context most representative to a plurality of users at a current time.

12. The method of claim 9, wherein the backstage rule is based on criteria that is likely to be of general interest to a plurality of users.

13. The method of claim 9, wherein each context is ranked in accordance with the number of higher rated individual items in such context.

14. The method of claim 9, wherein the backstage rule comprises a rule for determining which items are to be displayed on the screen based on higher individual rankings.

15. The method of claim 9, wherein the backstage rule comprises a combination of a rule which selects the most interesting items to the most users and a rule which selects the most recent items.