Granted Patent B2

US 12,744,125 · App. 18/826,555 · Granted Sep 22, 2026

Inventors: Russell Gould (Washington Crossing, PA); Soyoun Kristin Chung (Brooklyn, NY); Benjamin Serbiak (Georgetown, TX); Bryan Patrick Cunningham (Bridgewater, NJ); Jingting Zhang (Toronto, CA)

Assignee: Kenvue Brands LLC

G16H50/20G06T7/0012G06V10/762G16H10/20G16H20/10G16H50/30G16H50/70G06T2207/30201

Abstract

Claims (94)

1 . A computer-implemented method, comprising:

presenting, on a user interface (UI), a questionnaire;

receiving, on the UI, one or more responses to the questionnaire, the received one or more responses including a goal of a new user, the goal associated with a health outcome, wherein the health outcome is a stage of menopause;

generating a plurality of clusters of a plurality of existing profiles having a similarity above a similarity threshold, wherein generating the plurality of clusters includes (i) performing a hierarchical feature collapse and (ii) selecting top features by performing a principal component analysis (PCA) to group profile features together to determine strength of similarity between the profile features, and wherein the profile features include both behavioral features and biological features;

generating, for the new user, a profile;

associating the generated profile into a cluster of the plurality of clusters based on the similarity for the generated profile being above the similarity threshold;

determining a persona for the cluster of the plurality of clusters, wherein the determined persona is an artificial profile representing one or more data values of the generated profile in the cluster;

storing, in a specialized data structure, the cluster together with the determined persona for the cluster;

for the associated cluster, identifying, in the specialized data structure, a highest ranked intervention for the health outcome for the determined persona; and

generating, by a machine learning (ML) model, a recommendation for the new user based on the associated cluster and the determined persona, wherein the generated recommendation includes the identified highest ranked intervention that, when applied, has a likelihood of accomplishing the goal associated with addressing the health outcome.

2 . The computer-implemented method of claim 1 , wherein generating the plurality of clusters further comprises:

identifying the plurality of existing profiles, each existing profile of the plurality of existing profiles including a set of variables;

for each of the existing profiles, identifying a value for each variable of the set of variables;

determining, by a clustering algorithm, a first existing profile and a second existing profile, of the plurality of existing profiles, have a similarity above the similarity threshold; and

generating, by the clustering algorithm, the cluster including the first existing profile and the second existing profile.

3 . The computer-implemented method of claim 2 , wherein associating the generated profile into the cluster of the plurality of clusters further comprises:

identifying a new set of variables for the generated profile;

identifying a value for each variable of the new set of variables for the generated profile;

based on the identified value for each variable of the new set of variables for the generated profile, determining, by the clustering algorithm, the cluster of the plurality of clusters most similar to the generated profile; and

associating the generated profile into the determined cluster.

4 . The computer-implemented method of claim 3 , wherein the set of variables include variables related to at least one of age, number of symptoms, types of symptoms, severity of symptoms, ethnicity, income, geographical location, or awareness of menopause.

5 . The computer-implemented method of claim 1 , wherein generating the recommendation for the user further comprises:

determining the health outcome associated with the cluster of the plurality of clusters;

identifying the intervention that, when applied, has a likelihood of addressing the determined health outcome, wherein the identified intervention includes at least one of a treatment for a symptom of the stage of menopause or educational content associated with the stage of menopause; and

generating, by the ML model, the recommendation for the user, the recommendation including the intervention.

6 . The computer-implemented method of claim 1 , further comprising:

determining prior consent for named customer data is received;

capturing both anonymous data and the named customer data via user navigation and input context on the UI;

storing the captured anonymous data and the named customer data in the specialized data structure; and

training a second ML model to associate the generated profile for the new user into the cluster based on the captured anonymous data and the named customer data.

7 . The computer-implemented method of claim 1 , further comprising:

receiving feedback indicating a result of the recommendation; and

based on the received feedback, updating the ML model.

8 . The computer-implemented method of claim 1 , further comprising:

receiving updated information from the new user;

based on the received updated information, associating the generated profile into a second cluster of the plurality of clusters, the second cluster different than the cluster; and

generating, by the ML model, a second recommendation for the user based on the associated second cluster.

9 . An apparatus comprising:

a user interface (UI) configured to present a questionnaire;

a memory configured to store a specialized data structure; and

a processor coupled to the memory configured to:

receive, via the UI, one or more responses to the questionnaire, the received one or more responses including a goal of a new user, the goal associated with a health outcome, wherein the health outcome is a stage of menopause;

generate a plurality of clusters of a plurality of existing profiles having a similarity above a similarity threshold, wherein generating the plurality of clusters includes (i) performing a hierarchical feature collapse and (ii) selecting top features by performing a principal component analysis (PCA) to group profile features together to determine strength of similarity between the profile features, and wherein the profile features include both behavioral features and biological features;

generate, for the new user, a profile associated with the user based on the received responses to the questionnaire;

associate the generated profile into a cluster of a plurality of clusters based on the similarity for the generated profile being above the similarity threshold;

determine a persona for the cluster of the plurality of clusters, wherein the determined persona is an artificial profile representing one or more data values of the generated profile in the cluster;

store, in the specialized data structure, the cluster together with the determined persona for the cluster;

for the associated cluster, identifying, in the specialized data structure, a highest ranked intervention for the health outcome for the determined persona; and

execute a machine learning (ML) model to generate a recommendation for the new user based on the associated cluster and the determined persona, wherein the generated recommendation includes the identified highest ranked intervention that, when applied, has a likelihood of accomplishing the goal associated with addressing the health outcome.

10 . The apparatus of claim 9 , wherein the processor is further configured to:

identify the plurality of existing profiles, each existing profile of the plurality of existing profiles including a set of variables;

for each of the existing profiles, identify a value for each variable of the set of variables;

execute a clustering algorithm to determine a first existing profile and a second existing profile, of the plurality of existing profiles, have a similarity above the similarity threshold; and

generate, by the clustering algorithm, the cluster including the first existing profile and the second existing profile.

11 . The apparatus of claim 10 , wherein, to associate the generated profile into the cluster of the plurality of clusters, the processor is further configured to:

identify a new set of variables for the generated profile;

identify a value for each variable of the new set of variables for the generated profile;

based on the identified value for each variable of the new set of variables for the generated profile, execute the clustering algorithm to determine the cluster of the plurality of clusters most similar to the generated profile; and

associate the generated profile into the determined cluster.

12 . The apparatus of claim 11 , wherein the set of variables include variables related to at least one of age, number of symptoms, types of symptoms, severity of symptoms, ethnicity, income, geographical location, or awareness of menopause.

13 . The apparatus of claim 9 , wherein, to generate the recommendation for the user, the processor is further configured to:

determine the health outcome associated with the cluster of the plurality of clusters;

identify the intervention that, when applied, has a likelihood of addressing the determined health outcome; and

execute the ML model to generate the recommendation for the user, the recommendation including the intervention.

14 . The apparatus of claim 13 , wherein:

the identified intervention includes at least one of a treatment for a symptom of the stage of menopause or educational content associated with the stage of menopause.

15 . The apparatus of claim 9 , wherein the processor is further configured to:

receive updated information from the new user;

based on the received updated information, associate the generated profile into a second cluster of the plurality of clusters, the second cluster different than the cluster; and

execute the ML model to generate a second recommendation for the user based on the associated second cluster.

16 . One or more non-transitory computer readable media storing instructions that, when executed by a processor, cause the processor to:

generate a plurality of clusters, each cluster of the generated plurality of clusters including at least a stage of menopause and symptoms experienced related to menopause, wherein generating the plurality of clusters includes (i) performing a hierarchical feature collapse and (ii) selecting top features by performing a principal component analysis (PCA) to group profile features together to determine strength of similarity between the profile features, and wherein the profile features include both behavioral features and biological features;

determine a persona for each of the generated clusters, wherein the determined persona is an artificial profile representing one or more data values of a generated profile in the cluster;

store, in a specialized data structure, each of the generated clusters together with the determined persona for the cluster;

generate, for a new user, a profile based on one or more responses to a questionnaire, the generated profile including at least an identified user stage of menopause and user symptoms experienced related to menopause, wherein the one or more responses include a goal of the new user;

associate the generated profile into a cluster of the plurality of clusters; and

generate, by a machine learning (ML) model, a recommendation for the user based on the associated cluster and the determined persona, wherein the generated recommendation includes an intervention that, when applied, has a likelihood of accomplishing the goal associated with addressing the identified user stage of menopause.

17 . The one or more non-transitory computer readable media of claim 16 , further storing instructions for generating the recommendation for the user that, when executed by the processor, cause the processor to:

identify an intervention associated with the associated cluster that, when applied, has a likelihood of addressing the identified user stage of menopause, the identified intervention including at least one of a treatment for a symptom of the stage of menopause or educational content associated with the stage of menopause; and

generating, by the ML model, the recommendation for the user, the recommendation including the intervention.

18 . The one or more non-transitory computer readable media of claim 16 , further storing instructions for generating the plurality of clusters that, when executed by the processor, cause the processor to:

identify a plurality of existing profiles, each existing profile of the plurality of existing profiles including a set of variables;

for each of the existing profiles, identify a value for each variable of the set of variables;

determine, by a clustering algorithm, a first existing profile and a second existing profile, of the plurality of existing profiles, have a similarity above a similarity threshold; and

generate, by the clustering algorithm, the cluster including the first existing profile and the second existing profile.

19 . The one or more non-transitory computer readable media of claim 18 , further storing instructions for associating the generated profile into the cluster of the plurality of clusters that, when executed by the processor, cause the processor to:

identify a new set of variables for the generated profile;

identify a value for each variable of the new set of variables for the generated profile;

based on the identified value for each variable of the new set of variables for the generated profile, determine, by the clustering algorithm, the cluster of the plurality of clusters most similar to the generated profile; and

associate the generated profile into the determined cluster.

20 . The one or more non-transitory computer readable media of claim 16 , further storing instructions that, when executed by the processor, cause the processor to:

receive updated information from the new user;

based on the received updated information, associate the generated profile into a second cluster of the plurality of clusters, the second cluster different than the cluster; and

generate, by the ML model, a second recommendation for the user based on the associated second cluster.

Assignments (4)

Continuity (2)

References Cited (25)