Pangenetic web item recommendation system

• US 8,386,519 B2
• Filed: 12/30/2008
• Issued: 02/26/2013
• Est. Priority Date: 12/30/2008
• Status: Active Grant

• Current Assignee: 23andme Incorporated (23andMe Holding Co.)
• Original Assignee: Expanse Networks, Inc.
• Inventors: Kenedy, Andrew Alexander, Eldering, Charles Anthony
• Primary Examiner(s): LEWIS, CHERYL RENEA
• Application Number:
  US12/346,707
• Publication Number:
  US 20100169340A1
• Time in Patent Office: 1,519 Days
• Field of Search: 707/609, 707/705, 707/767, 707/784, 705/7.29
• US Class Current: 707/784
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0631 Item recommendations
• CPC Class Codes: G16H 10/60 for patient-specific data, e.g. for electronic patient records

Abstract

Computer based systems, methods, software and databases are presented in which correlations between web item preferences and pangenetic (genetic and epigenetic) attributes of individuals are used for pangenetic based web item recommendation in which a user can request and receive personalized online recommendations of web items that are based on the user'"'"'s pangenetic makeup. Data masking can be used to maintain privacy of sensitive portions of the pangenetic data.

Claims

1. A computer based method for online recommendation of items for a user, comprising:

a) receiving at least one item preference associated with the user along with one or more non-pangenetic attributes associated with the user;

b) accessing pangenetic data associated with the user wherein the pangenetic data associated with the user constitutes a pangenetic profile of the user, wherein said accessing step is in accord with a previously set user identification and authorization protocol whereby said pangenetic profile is associated with said user and is masked according to a predetermined data mask;

c) accessing a dataset containing item preferences of individuals who also share the at least one item preference associated with the user, wherein pangenetic and non-pangenetic data of the individuals are correlated with the item preferences and wherein the pangenetic data correlated with the item preferences are combinations of pangenetic data selected from pangenetic profiles of the individuals, and further wherein the one or more non-pangenetic attributes associated with the user match one or more non-pangenetic attributes associated with the individuals, wherein the correlations between the pangenetic data and the item preferences contained in the dataset are previously determined based on statistical associations between item preferences and pangenetic data associated with the individuals;

d) determining for each item preference, the quantity of matches between the pangenetic data associated with that item preference and the pangenetic data associated with the user; and

e) transmitting as output, based on the quantity of matches determined for each item preference, a listing of at least a portion of the item preferences to indicate recommended items for the user wherein the portion of the item preferences transmitted as output is determined by a predetermined threshold applied to the quantity of matches determined for each item preference.

2. The computer based method of claim 1, wherein the transmitting in step (e) is to at least one destination selected from the group consisting of the user, a database, a dataset, a computer readable memory, a computer readable medium, a computer processor, a computer network, a printout device, a visual display, and a wireless receiver.

3. The computer based method of claim 1, further comprising transmitting as output at least a portion of the item preferences to indicate non-recommended items for the user.

4. The computer based method of claim 1, wherein the pangenetic data correlated with the item preferences are pangenetic metadata.

5. The computer based method of claim 1, wherein the item preferences are ratings that indicate levels of satisfaction with the items indicated by the item preferences.

6. The computer based method of claim 5, wherein the ratings are average ratings of the items by the individuals.

7. The computer based method of claim 1, further comprising the steps of i) transmitting an authorization request for access to the pangenetic data associated with the user, and ii) receiving an authorization granting access to the pangenetic data associated with the user.

8. The computer based method of claim 1, wherein the receiving in step (a) is from a mobile computing device, and wherein the transmitting in step (e) is to the mobile computing device.

9. The computer based method of claim 1, wherein the listing is a rank listing, and wherein the rank of each item preference in the rank listing is based on the quantity of matches determined for each item preference.

10. The computer based method of claim 9, wherein the portion of the item preferences transmitted as output consists of item preferences having a rank within a range defined by at least one predetermined threshold applied to rank.

11. The computer based method of claim 9, wherein the rank of each item preference represented in the rank listing is determined by a score computed for each item preference based on the quantity of matches determined for each item preference.

12. The computer based method of claim 11, wherein the score for each item preference is computed using a quantitative similarity measure applied to the pangenetic data.

13. The computer based method of claim 1, wherein the correlations between the pangenetic data and the item preferences contained in the dataset are determined by computing statistical associations between pangenetic data of individuals and online behaviors which indicate the item preferences of the individuals.

14. The computer based method of claim 1, wherein the dataset is an item feedback matrix.

15. The computer based method of claim 1, wherein the accessing in step (b) is in accordance with an applied data mask, the method further comprising:

i) transmitting an authorization request for access to the pangenetic data associated with the user;

ii) receiving an authorization which grants access to the pangenetic data;

iii) accessing a data mask, wherein the data mask'"'"'s parameters are associated with the authorization; and

iv) applying the data mask to the pangenetic data.

16. The computer based method of claim 1, wherein the dataset comprises data records containing the item preferences of the individuals, the method further comprising:

i) identifying one or more clusters of data records, wherein within each cluster the data records share a similar pattern of item preferences as determined by a quantitative similarity measure;

ii) determining, by statistical association, pangenetic data that correlate with each of the one or more clusters; and

iii) identifying, by using a quantitative similarity measure, the cluster having the highest pangenetic similarity to the user to provide the portion of the item preferences to be transmitted as output in step (e).

17. The computer based method of claim 16, wherein the item preferences of the identified cluster comprise item rating values that are averaged prior to transmission as output.

18. The computer based method of claim 16, wherein the item preferences identified for transmission as output are a subset of item preferences selected from the identified cluster based on an item category relationship with the at least one item preference associated with the user.

19. The computer based method of claim 1, wherein the dataset comprises data records containing the item preferences of the individuals, and wherein the item preferences comprise item rating values, the method further comprising:

i) identifying one or more clusters of data records, wherein within each cluster the data records share a similar pattern of item preferences as determined by a quantitative similarity measure;

ii) determining, by statistical association, pangenetic data that correlate with each of the one or more clusters;

iii) identifying, by using a quantitative similarity measure, the cluster having the highest pangenetic similarity to the user; and

iv) identifying, by using a quantitative similarity measure within the cluster having the highest pangenetic similarity to the user, a subcluster of data records having the most similar pattern of item preferences to the user to provide the portion of the item preferences to be transmitted as output in step (e).

20. The computer based method of claim 19, wherein the item preferences of the identified subcluster comprise item rating values that are averaged prior to transmission as output.

21. The computer based method of claim 1, further comprising receiving feedback from the user with respect to one or more of the item preferences contained in the listing transmitted as output.

22. A program storage device readable by a machine and containing a set of instructions which, when read by the machine, causes execution of a computer based method for online recommendation of items for a user, comprising:

a) receiving at least one item preference associated with the user along with one or more non-pangenetic attributes associated with the user;

b) accessing pangenetic data associated with the user wherein the pangenetic data associated with the user constitutes a pangenetic profile of the user, wherein said accessing step is in accord with a previously set user identification and authorization protocol whereby said pangenetic profile is associated with said user and is masked according to a predetermined data mask;

c) accessing a dataset containing item preferences of individuals who also share the at least one item preference associated with the user, wherein pangenetic and non-pangenetic data of the individuals are correlated with the item preferences and wherein the pangenetic data correlated with the item preferences are combinations of pangenetic data selected from pangenetic profiles of the individuals, and further wherein the one or more non-pangenetic attributes associated with the user match one or more non-pangenetic attributes associated with the individuals, wherein the correlations between the pangenetic data and the item preferences contained in the dataset are previously determined based on statistical associations between item preferences and pangenetic data associated with the individuals;

d) determining for each item preference, the quantity of matches between the pangenetic data correlated with that item preference and the pangenetic data associated with the user; and

e) transmitting as output, based on the quantity of matches determined for each item preference, a listing of at least a portion of the item preferences to indicate recommended items for the user wherein the portion of the item preferences transmitted as output is determined by a predetermined threshold applied to the quantity of matches determined for each item preference.

23. The program storage device of claim 22, wherein the transmitting in step (e) is to at least one destination selected from the group consisting of the user, a database, a dataset, a computer readable memory, a computer readable medium, a computer processor, a computer network, a printout device, a visual display, and a wireless receiver.

24. A computer database system for online recommendation of items for a user, comprising:

a) a memory containing;

i) a first data structure containing pangenetic data associated with the user wherein the pangenetic data associated with the user constitutes a pangenetic profile of the user;

ii) a second data structure containing item preferences of individuals who also share at least one item preference associated with the user, wherein pangenetic data and non-pangenetic of the individuals are correlated with the item preferences;

b) a processor for;

i) receiving the at least one item preference associated with the user along with one or more non-pangenetic attributes associated with said user;

ii) accessing the first data structure, wherein said accessing step is in accord with a previously set user identification and authorization protocol whereby said pangenetic profile is associated with said user and is masked according to a predetermined data mask;

iii) accessing the second data structure;

iv) determining for each item preference, the quantity of matches between the pangenetic data correlated with that item preference and the pangenetic data associated with the user and wherein the pangenetic data correlated with the item preferences are combinations of pangenetic data selected from pangenetic profiles of the individuals, and further wherein the one or more non-pangenetic attributes associated with the user match one or more non-pangenetic attributes associated with the individuals, wherein the correlations between the pangenetic data and the item preferences contained in the dataset are previously determined based on statistical associations between item preferences and pangenetic data associated with the individuals; and

v) transmitting as output, based on the quantity of matches determined for each item preference, a listing of at least a portion of the item preferences to indicate recommended items for the user wherein the portion of the item preferences transmitted as output in step is determined by a predetermined threshold applied to the quantity of matches determined for each item preference.

25. The computer database system of claim 24, wherein the transmitting in part (v) is to at least one destination selected from the group consisting of the user, a database, a dataset, a computer readable memory, a computer readable medium, a computer processor, a computer network, a printout device, a visual display, and a wireless receiver.