Recommendation networks for ranking recommendations using trust rating for user-defined topics and recommendation rating for recommendation sources

• US 7,761,399 B2
• Filed: 08/21/2006
• Issued: 07/20/2010
• Est. Priority Date: 08/19/2005
• Status: Active Grant

• Current Assignee: SitNet LLC
• Original Assignee: Evree, LLC
• Inventors: Evans, Matthew Reid
• Primary Examiner(s): Sparks; Donald
• Assistant Examiner(s): CHANG, LI WU
• Application Number:
  US11/507,699
• Publication Number:
  US 20070064626A1
• Time in Patent Office: 1,429 Days
• Field of Search: 706/55
• US Class Current: 706/55
• CPC Class Codes: G06F 16/958 Organisation or management of web site content, e.g. publishing, maintaining pages or automatic linking
• CPC Class Codes: G06Q 10/10 Office automation; Time management

Abstract

A recommendation network is described. In some embodiments, the recommendation network includes recommenders that explicitly or implicitly recommend, rate or refer items and recommendation receivers that receive the recommendations. In some embodiments, the recommenders can be recommendation receivers, and vice versa. In some embodiments, recommendation receivers assign trust ratings to recommenders. The recommendation receiver can assign separate trust ratings to individual topics for which the recommendation receiver trusts the recommender. The separate trust ratings represent the recommendation receiver'"'"'s amount of trust in the recommender to makes valuable recommendations for the specific topic. The recommendation network can use the separate trust ratings, along with ratings provided by the recommender, to rank recommendations per the separate topics. The recommendation receiver can assign the recommender to different bundles, topics, channels, etc. to which other recommendation receivers can subscribe.

Claims

1. A system, comprising:

a server configured to

generate, in response to a first input provided by a user source, a recommendation container configured to receive content recommendations from at least one recommender source that is within a social network of the user source, and wherein the recommendation container is user definable by the user source,

assign a user-defined topic to the recommendation container, wherein the user-defined topic is defined by a second input from the user source,

associate, in response to a third input from the user source, the at least one recommender source with the recommendation container,

assign a trust rating in response to a fourth input from the user source, to the at least one recommender source specifically for the user-defined topic, wherein the trust rating represents a degree of trust that the user source has in the at least one recommender source, as a recommendation source, to provide content recommendations of value specifically for the user-defined topic,

detect a set of content recommendations provided by the at least one recommender source, wherein the set of content recommendations are on a plurality of topics, and wherein at least one of the plurality of topics includes the user-defined topic;

filter the set of content recommendations using the user-defined topic forming a subset of content recommendations that are classified by the user-defined topic,

determine individual recommendation ratings for each content recommendation in the subset of content recommendations, wherein the individual recommendation ratings are set by the at least one recommender source and represent individual degrees of preference that the at least one recommender source has for each content recommendation in the subset of content recommendations

calculate individual ranking scores for each content recommendation in the subset of content recommendations by computing each individual recommendation rating with the trust rating, and

rank the subset of content recommendations into a ranked list based on the individual ranking scores for each content recommendation in the subset of content recommendations; and

a client device configured to

receive the first, second, third, and fourth inputs from the user source,

provide the first, second, third, and fourth inputs to the server,

receive the ranked list from the server, and

present the ranked list on a display associated with the client device.

2. The system of claim 1, wherein the server is further configured to

determine topic descriptions assigned by the recommender source to the content recommendations,

determine that a subset of the topic descriptions, for only the subset of content recommendations, include topic descriptions that match the user-defined topic, and

include in the ranked list only the subset of content recommendations.

3. The system of claim 1, wherein the server is further configured to

receive an additional trust rating, via sixth input by the user source, assigned specifically to the recommendation container wherein the additional trust rating indicates a degree of trust that the user source has in the recommendation container, as an additional recommendation source, to provide recommendations of value on the user-defined topic, and

recalculate the individual ranking scores for each content recommendation in the subset of content recommendations by computing each individual recommendation rating with the trust rating and with the additional trust rating.

4. The system of claim 1, wherein the server is further configured to

provide an instance of the recommendation container to at least one additional user source,

receive at least one additional trust rating from the at least one additional user source, wherein the at least one additional trust rating represents a degree of trust that the at least one additional user source has in the recommendation container, as an additional recommendation source, specifically for the user-defined topic, and

rank the subset of content recommendations in at least by computing the individual ranking scores with the at least one additional trust rating.

5. The system of claim 1, where the server is further configured to

receive a recommendation temporality rating, wherein the recommendation temporality rating indicates a user-defined, timeliness of relevance for the subset of content recommendations, wherein the timeliness of the relevance is indicative of a user'"'"'s perceived degree that content is relevant over time, and

assign the recommendation temporality rating to at least one of the subset of content recommendations, and

rank the subset of content recommendations into the ranked list using the individual ranking scores and the recommendation temporality rating.

6. The system of claim 1, wherein the at least one recommendation source represents a subset of social-network user entities linked to a website account for the user source.

7. The system of claim 1, where the server is further configured to

provide an instance of the recommendation container to at least one additional user source,

channel the subset of content recommendations into the instance of the recommendation container,

determine at least one additional user-defined topic indicated by the at least one additional user source,

assign the at least one additional user-defined topic to the instance of the recommendation container, and

filter the subset of content recommendations channeled into the instance of the recommendation container using the additional user-defined topic.

8. A computer-implemented method, wherein one or more processors for a computer associated with a computerized social network perform operations comprising:

generating a first recommendation container for a first user source, wherein the first commendation container is user-definable by the first user source;

presenting to the first user source, in a graphical user interface, a listing of a plurality of additional recommendation containers that belong to a plurality of additional user sources on a computerized social network;

determining a selection, by the first user source via the graphical user interface, of a second recommendation container from the listing of the plurality of additional recommendation containers, wherein the second recommendation container belongs to a second user source and is user-definable by the second user source and not the first user source, and wherein the second recommendation container includes content recommendations that are categorized by a set of user-defined topics specified by the second user source;

linking the second recommendation container with the first recommendation container, wherein said linking enables content recommendations from the second recommendation container to flow into the first recommendation container;

determining a selection by the first user source of one user-defined topic from the multiple user-defined topics;

filtering the content recommendations, using the one user-defined topic, causing only a subset of the content recommendations that are classified by the one of user-defined topic to flow into the first recommendation container from the second recommendation container;

assigning a trust rating value, specified by the first user source, to the second recommendation container specifically for the one user-defined topic, wherein the trust rating value represents a degree of trust that the first user source has in the second user source to provide noteworthy content recommendations, through the second recommendation container, specifically for the one user-defined topic;

determining a recommendation rating value assigned by the second user source to at least one content recommendation from the subset of content recommendations, wherein the recommendation rating value indicates a degree of preference that the second user source has for the at least one content recommendation;

calculating a ranking score for the at least one content recommendation by combining the recommendation rating value with the trust rating value; and

ranking the at least one content recommendation in a ranked list based on the ranking score.

9. The computer-implemented method of claim 8, wherein determining the selection by the first user source of one user-defined topic from the multiple user-defined topics includes,

presenting in the graphical user interface, responsive to the selection by the first user source of the second recommendation container, a listing of the multiple user-defined topics that are associated with the second recommendation container, and

determining the selection by the first user source of the one user-defined topic from the listing of the multiple user-defined topics.

10. The computer-implemented method of claim 8, wherein calculating the ranking score for the at least one content recommendation further comprises,

determining at least one additional recommendation container linked to the second recommendation container,

determining at least one additional rating score for the at least one additional recommendation container, and

computing the ranking score by combining the recommendation rating value with the trust rating value and the at least one additional rating score.

11. The computer-implemented method of claim 8, further comprising:

receiving search criteria from the first user source, before presenting to the first user source the listing of the plurality of additional recommendation containers, wherein the search criteria includes a textual description substantially similar to a description of the one user-defined topic;

searching, using the search criteria, textual characteristics of content recommendations within the computerized social network by comparing at least some portion of the textual description to the textual characteristics of the content recommendations; and

finding, based on the searching, the plurality of additional recommendation containers, wherein each of the plurality of additional recommendation containers includes some content recommendations with textual characteristics that match the at least some portion of the textual description.

12. The computer-implemented method of claim 8 further comprising:

determining a selection of the first recommendation container by a third user source;

linking the first recommendation container to an additional recommendation container belonging to the third user source causing the at least one content recommendation and the ranking score to flow into the additional recommendation container;

assigning an additional trust rating value, indicated by the third user source, to the first recommendation container, wherein the additional trust rating value represents a degree of trust that the third user source has in the first user source to provide noteworthy content recommendations, through the first recommendation container, specifically for the one user-defined topic;

calculating an additional ranking score for the at least one content recommendation by combing the ranking score with the additional trust rating value; and

ranking the at least one content recommendation within the additional recommendation container using the additional ranking score.

13. The computer-implemented method of claim 12 further comprising:

determining at least one content blocking filter assigned by the third user source to the first recommendation container, and

blocking a portion of the subset of content recommendations from flowing into the additional recommendation container based on the content blocking filter.

14. The computer-implemented method of claim 12, further comprising:

including an additional content recommendation in the first recommendation container, wherein the additional content recommendation is made by the first user source and classified by the first user source with the one user-defined topic;

assigning an additional recommendation rating, made by the first user source, to the additional content recommendation;

causing the additional content recommendation and the additional recommendation rating to flow into the additional recommendation container;

calculating a second additional ranking score for the additional content recommendation by combing the ranking score, the additional ranking score, and the additional recommendation rating; and

ranking the additional content recommendation within the additional recommendation container using the second additional ranking score.

15. The computer-implemented method of claim 8, further comprising:

assigning a user-defined subtopic to the one user-defined topic, wherein the user-defined sub-topic is specified by the first user source;

assigning an additional content recommendation by the first user source to the first content container for the user-defined sub-topic;

assigning an additional recommendation rating value, specified by the first user source, to the additional content recommendation;

calculating an additional ranking score for the additional content recommendation using the trust rating value and the additional recommendation rating value; and

ranking the additional content recommendation in the first content container based on the additional ranking score.

16. The computer-implemented method of claim 8, further comprising:

receiving a recommendation temporality rating by the first user source, wherein the recommendation temporality rating indicates a perceived degree of relevance over time of the least one content recommendation;

assigning the recommendation temporality rating to the least one content recommendation; and

recalculating the ranking score using the recommendation temporality rating.

17. One or more non-transitory computer readable media having instructions stored thereon, which when executed by a set of one or more processors causes the set of one or more processors to perform operations comprising:

generating, in response to a first input provided by a user source, a recommendation container configured to receive content recommendations from at least one recommender source that is a subset of social-network entities linked to the user source, wherein the recommendation container is user definable by only the user source;

assigning a user-defined topic to the recommendation container, wherein the user-defined topic is defined by a second input from the user source;

associating, in response to a third input from the user source, the at least one recommender source with the recommendation container;

assigning a trust rating, in response to a fourth input from the user source, to the at least one recommender source specifically for the user-defined topic, wherein the trust rating-represents a degree of trust that the user source has in the at least one recommender source, as a recommendation source, to provide content recommendations of value specifically for the user-defined topic;

detecting a set of content recommendations provided by the at least one recommender source, wherein the set of content recommendations are on a plurality of topics, and wherein at least one of the plurality of topics includes the user-defined topic;

filtering the set of content recommendations using the user-defined topic forming a subset of content recommendations that are classified by the user-defined topic;

determining individual recommendation ratings for each content recommendation in the subset of content recommendations, wherein the individual recommendation ratings are set by the at least one recommender source and represent individual degrees of preference that the at least one recommender source has for each content recommendation in the subset of content recommendations;

calculating individual ranking scores for each content recommendation in the subset of content recommendations by computing each individual recommendation rating with the trust rating; and

ranking the subset of content recommendations into a ranked list based on the individual ranking scores for each content recommendation in the subset of content recommendations.

18. The one or more non-transitory computer readable media of claim 17, the operations further comprising:

determining topic descriptions assigned by the recommender source to the content recommendations;

determining that a subset of the topic descriptions, for only the subset of content recommendations, include topic descriptions that match the user-defined topic; and

including in the ranked list only the subset of content recommendations.

19. The one or more non-transitory computer readable media of claim 17, the operations further comprising:

receiving an additional trust rating, via sixth input by the user source, assigned specifically to the recommendation container, wherein the additional trust rating indicates a degree of trust that the user source has in the recommendation container, as an additional recommendation source, to provide recommendations of value on the user-defined topic; and

recalculating the individual ranking scores for each content recommendation in the subset of content recommendations by computing each individual recommendation rating with the trust rating and with the additional trust rating.

20. The one or more non-transitory computer readable media of claim 17, further comprising:

providing an instance of the recommendation container to at least one additional user source;

receiving at least one additional trust rating from the at least one additional user source, wherein the at least one additional trust rating represents a degree of trust that the at least one additional user source has in the recommendation container, as an additional recommendation source, specifically for the user-defined topic; and

ranking the subset of content recommendations into at least one additional ranked list by computing the individual ranking scores with the at least one additional trust rating.