Recommender system for identifying a new set of media items responsive to an input set of media items and knowledge base metrics

• US 7,734,569 B2
• Filed: 02/03/2006
• Issued: 06/08/2010
• Est. Priority Date: 02/03/2005
• Status: Expired due to Fees

• Current Assignee: Apple Inc.
• Original Assignee: Strands Incorporated (CRIBIS Holding SRL)
• Inventors: Martin, Francisco J., Shur, Jim, Torrens, Marc
• Primary Examiner(s): Wu; Yicun
• Assistant Examiner(s): LAWSON, DONALD L
• Application Number:
  US11/346,818
• Publication Number:
  US 20060184558A1
• Time in Patent Office: 1,586 Days
• Field of Search: 707/5, 707/200, 707/104.1, 345/716, 715/500.1, 709/224
• US Class Current: 1/1
• CPC Class Codes: G06F 16/435 Filtering based on additional data, e.g. user or group profiles
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0631 Item recommendations
• CPC Class Codes: H04N 21/466 Learning process for intelligent management, e.g. learning user preferences for recommending movies services using the results of monitoring in broadcast systems H04H60/61
• CPC Class Codes: H04N 21/4668 for recommending content, e.g. movies
• CPC Class Codes: H04N 21/4825 using a list of items to be played back in a given order, e.g. playlists
• CPC Class Codes: H04N 7/163 by receiver means only

Abstract

Systems and methods are disclosed for identifying a new set of media items in response to an input set (or “query set”) of media items and knowledge base metrics. The system uses a knowledge base consisting of a collection of mediasets. Various metrics among media items are considered by analyzing how the media items are grouped to form the mediasets in the knowledge base. Such association or “similarity” metrics are preferably stored in a matrix form that allows the system to efficiently identify a new set of media items that complements the input set of media items.

Claims

1. A computer-implemented method comprising:

accessing a knowledge base stored in a memory, the knowledge base including a plurality of collected mediasets;

wherein each collected mediaset includes a plurality of entries grouped together according to preferences of a user; and

wherein each of the plurality of entries is configured to identify a corresponding media item without including content of the corresponding media item;

configuring a processor to;

process the knowledge base by;

defining every pair of entries in each of the collected mediasets; and

computing a correlation metric for every defined pair of entries in each of the collected mediasets;

receive at least one input that identifies at least one input media item without including content corresponding to the at least one input media item;

identify at least one of the defined pairs of entries in which one entry of the at least one of the defined pairs of entries corresponds to the at least one input;

select at least one candidate entry among the identified at least one of the defined pairs of entries by evaluating the corresponding computed correlation metrics;

wherein each candidate entry is an entry of the identified at least one of the defined pairs; and

wherein the at least one candidate entry identifies a corresponding candidate output media item to form an output set; and

cause transmission of the output set over a communication channel.

2. The computer-implemented method according to claim 1 wherein the input comprises first meta data identifying a first song.

3. The computer-implemented method according to claim 2 wherein the output set comprises second meta data identifying a second song.

4. The computer-implemented method according to claim 1 wherein the computer-implemented method comprises further configuring the processor to:

compute a co-concurrency metric configured to indicate a number of entries within a same collected mediaset in which a first entry corresponding to the at least one input appears together with a second entry; and

select a candidate output entry responsive to the co-concurrency metric exceeding a predetermined threshold metric value.

5. The computer-implemented method according to claim 1

wherein the plurality of collected mediasets are ordered;

and

wherein the computer-implemented method comprises further configuring the processor to;

compute a pre-concurrency metric configured to indicate a number of entries within a same collected mediaset in which a first entry corresponding to the at least one input occurs before a second entry of the pair; and

select a candidate output entry responsive to the pre-concurrency metric exceeding a predetermined threshold metric value.

6. The computer-implemented method according to claim 1 comprising further configuring the processor to:

apply a predetermined selection criterion to the plurality of entries in the plurality of collected mediasets from which the at least one candidate entry was selected.

7. The computer-implemented method according to claim 1 wherein:

the at least one input comprises first meta data configured to identify a first song;

the knowledge base comprises a plurality of collected song mediasets; and

the at least one candidate entry comprises second meta data configured to identify a second song.

8. The computer-implemented method according to claim 7 comprising further configuring the processor to cause transmission or display of the second meta data over a channel.

9. The computer-implemented method according to claim 7 comprising further configuring the processor to filter the output set with a predetermined additional external constraint.

10. The computer-implemented method according to claim 9 comprising further configuring the processor to maximize the predetermined additional external constraint.

11. The computer-implemented method according to claim 9 comprising further configuring the processor to apply a predetermined preference or criterion to the at least one candidate entry.

21. The computer-implemented method according to claim 1 wherein:

the mediasets are ordered;

the computing includes computing a post-concurrency metric for each input media item relative to every pair of entries in a mediaset in which a first entry identifying the input media item occurs after a second entry of the pair; and

the selecting includes selecting as a candidate output media item each media item corresponding to an entry for which the post-concurrency metric exceeds a predetermined threshold metric value.

12. A computer-implemented method for dynamically generating an output set responsive to an input set comprising:

storing a plurality of collected mediasets in a memory;

wherein each collected mediaset includes a plurality of entries grouped together according to preferences of a user; and

wherein each of the plurality of entries corresponds to a collected media item without including content of the collected media item;

configuring a processor to;

receive the input set including an input entry identifying a corresponding at least one input media item;

define every pair of entries in each of the collected mediasets;

compute a metric for every defined pair of entries in each of the collected mediasets;

identify, from among the defined pairs of entries, an entry in each defined pair of entries that corresponds to the input entry; and

select among the identified entries, a candidate entry that identifies a corresponding candidate media item having a highest metric, the highest metric being configured to indicate an affinity between the candidate entry and the input entry.

13. The computer-implemented method according to claim 12 wherein the metric comprises one of pre-concurrency metric, a post-concurrency metric, or a co-concurrency metric.

14. The computer-implemented method according to claim 12 wherein:

the input set includes first meta data identifying first songs;

the plurality of collected mediasets includes a plurality of collected mediasets of songs; and

the output set includes second meta data identifying second songs.

15. The computer-implemented method according to claim 14 further comprising:

repeating the selecting among the identified entries for other candidate entries to form a set of candidate entries; and

filtering the set of candidate entries prior to transmission.

16. The computer-implemented method according to claim 15 wherein the filtering includes limiting the set of candidate entries to a determined type of items.

17. The computer-implemented method according to claim 15 wherein the filtering includes limiting the set of candidate entries to items that meet defined keyword-based meta data criteria, thereby personalizing the recommendations.

18. A computer-implemented method, comprising:

storing a knowledge base in a memory, the knowledge base including a plurality of mediasets, each mediaset including a plurality of entries being grouped together according to preferences of a user and the plurality of entries identifying a corresponding plurality of media items without including content of the corresponding plurality of media items; and

configuring a processor to;

for each entry corresponding to a media item in each of the plurality of mediasets, identify every other entry corresponding to the media item in a same mediaset, so as to define every pair of entries in each of the plurality of mediasets; and

compute a metric for every defined pair of entries that identifies every pair of media items in the knowledge base;

receive at least one input corresponding to at least one input media item;

identify, from among the defined pairs of entries, an entry in each defined pair of entries that corresponds to the input entry;

select among the identified entries, a candidate entry that identifies a corresponding candidate media item having a highest metric, the highest metric being configured to indicate an affinity between the candidate entry and the input entry.

19. The computer-implemented method according to claim 18 further comprising configuring the processor to:

for each pair of media items, adjust a concurrency metric value responsive to each additional occurrence of a same pair of media items in the mediasets, wherein concurrency metric value is a co-concurrency metric value; and

increment the co-concurrency metric value responsive to every additional occurrence of the same pair of media items in the mediasets.

20. The computer-implemented method according to claim 18 wherein:

the plurality of entries for the plurality of media items are ordered within each of the mediasets in the knowledge base;

configuring the processor to adjust includes configuring the processor to define ordered pairs of media items that appear in any of the mediasets in the knowledge base in a same order;

the concurrency metric value is an ordered concurrency metric value; and

configuring the processor to adjust includes configuring the processor to increment the ordered concurrency metric value responsive to every additional occurrence of the same pairs of media items in the same order in the knowledge base.