Online Clothing E-commerce Systems and Methods with Machine-learning Based Sizing Recommendation

Content Provider	The Lens
Abstract	Methods, systems, and storage media for providing sizing information is described. In embodiments, a computing device may obtain purchase information associated with individual users. Each of the individual users may correspond to at least one user device of a plurality of user devices. The purchase information may indicate a purchase of an item by the individual users. The computing device may obtain feedback information associated with the item that is provided by a subset of the individual users. The computing device may generate sizing information for the item based on the purchase information and the feedback information. The computing device may generate a recommendation for the item. Other embodiments may be described and/or claimed.
Related Links	https://www.lens.org/lens/patent/012-502-083-957-225/frontpage
Language	English
Publisher Date	2019-12-03
Access Restriction	Open
Content Type	Text
Resource Type	Patent
Jurisdiction	United States of America
Date Applied	2015-09-24
Agent	Schwabe, Williamson & Wyatt, P.c.
Applicant	Intel Corp
Application No.	201514864560
Claim	A computing system for providing online clothing recommendation service, the computing system comprising: processor circuitry arranged to operate a sizing information module to: identify purchase information entries in a purchase information database (DB) associated with an item and feedback information entries in a feedback information DB associated with the item, wherein the purchase information is associated with individual users that have purchased the item and each of the individual users corresponds to at least one user device of a plurality of user devices, and the feedback information is information that is indicative of user interactions with the item provided by a subset of the individual users via respective user devices, classify, as belonging to a first cluster, feedback information entries of the identified feedback information entries determined to be related to one or more characteristics of the item determined from the feedback information entries, classify, as belonging to a second cluster, feedback information entries of the identified feedback information entries determined to not be related to the one or more characteristics, generate a machine learning (ML) model that clusters user information of the individual users as belonging to one of the first cluster or the second cluster, and store, as sizing information in a sizing DB, the ML model in association with body measurement data indicated by the purchase information entries and preference information associated with the item as indicated by the feedback information entries; and the processor circuitry is arranged to operate a recommendation module to: input first user data of a first user into the ML model to determine whether the first user should be classified as belonging to the first cluster or the second cluster, and generate a recommendation message for a first user device associated with the first user based on a cluster to which the first user is classified according to the ML model; and communication circuitry, communicatively coupled with the processor circuitry, the communication circuitry arranged to obtain the first user data, and provide the recommendation message to the first user device. The computing system of claim 1 , wherein the first user data includes at least one of first body measurements associated with the first user, first demographic information associated with the first user, or first preference information of the first user. The computing system of claim 2 , wherein the first user data is to be obtained via an online form of a webpage provided by a merchant service or obtained from data associated with one or more applications stored on a computer-readable medium of the first user device associated with the first user, wherein the one or more applications include at least one of a web browser, an application for purchasing items via the merchant service, or an application associated with a third party communication platform; or wherein at least the first body measurements are to be obtained by at least one sensor associated with the first user device. The computing system of claim 1 , wherein purchase information in the purchase information entries includes body measurement information to indicate at least one body measurement for a corresponding one of the individual users or preference information to indicate at least one preference for the corresponding ones of the individual users; and feedback information in the feedback information entries includes the body measurement information of the subset of the individual users, the preference information of the subset of the individual users, and return information of the subset of the individual users, wherein the return information is to indicate a reason why the item was returned to a merchant service, and wherein the sizing information module is to receive the return information from a return DB associated with the merchant service. The computing system of claim 4 , wherein the first cluster at least represents characteristics of items that were returned and the reason indicated by the return information is representative of a size related return, and the second cluster at least represents characteristics of the items purchased that are not in the first cluster, and wherein a clustering operation to generate the first cluster and the second cluster is based on the body measurement information and the preference information of the individual users, and wherein, to generate the sizing information, the sizing information module is to: store, in the sizing DB, the model in association with the body measurement information and the preference information. The computing system of claim 5 , wherein to generate the recommendation message for the first user, the recommendation module is arranged to: determine, based on the first user data, a product of interest, first body measurements associated with the first user, and first preference information of the first user; obtain the model from the sizing DB when the product of interest is determined to have one or more characteristics in common with the one or more characteristics of the item; determine first user characteristics associated with the user of the first user device based on the first preference information; determine, using the ML model and based on at least one first user characteristic of the first user characteristics and the first body measurements, whether the first user characteristics should be classified as belonging to the first cluster or the second cluster; and generate the recommendation message to recommend the item for purchase when the first user characteristics are determined to be classified as belonging to the second cluster, or generate the recommendation message to not recommend the item for purchase when the first user characteristics are is-determined to be classified as belonging to the first cluster. The computing system of claim 6 , wherein the ML model is based on a minimum body measurement value and a maximum body measurement value from the body measurement information of individual users associated with the purchase information and not among the subset of the individual users associated with the feedback information entries, and the recommendation module is arranged to use the ML model to classify the first user data as belonging to the second cluster when the first user data includes a body measurement within the minimum body measurement value and the maximum body measurement value. The computing system of claim 6 , wherein the ML model is a support vector machine (SVM) model, and the sizing information module is arranged to use the purchase information entries and the feedback information entries as training data to generate the ML mode The computing system of claim 8 , wherein the recommendation message to not recommend the item for purchase includes a recommendation for another item to purchase based on the first user characteristics. The computing system of claim 1 , wherein the communication circuitry is arranged to receive a request message, and the recommendation module is arranged to generate the recommendation in response to receipt of the request message. The computing system of claim 10 , wherein the communication circuitry is arranged to receive the request message from a merchant application associated with a merchant service that is to sell the item, and wherein the request message is based on text entered into the merchant application, or the request message is based on a bar code or a quick response code obtained using at least one sensor of the first user device via the merchant application. The computing system of claim 11 , wherein the first user data includes position information associated with the first user device, and the recommendation module is arranged to generate the recommendation based on a determination as to whether the distance between the item and the first user device is determined to be within a predefined distance. At least one non-transitory computer-readable medium (NTCRM) that comprises instructions, wherein execution of the instructions by a computing device is to cause the computing device to: obtain purchase information associated with individual users each of which corresponds to at least one user device of a plurality of user devices, wherein the purchase information is to indicate a purchase of an item by the individual users; obtain feedback information associated with the item wherein the feedback information is to be provided by a subset of the individual users; classify, as belonging to a first cluster, feedback information determined to be related to one or more characteristics of the item as determined from the feedback information; classify, as belonging to a second cluster, feedback information determined to not be related to the one or more characteristics; generate a machine learning (ML) model that clusters user information of the individual users as belonging to one of the first cluster or the second cluster; store, as sizing information in a sizing database (DB), the ML model in association with body measurement data indicated by purchase information entries and preference information associated with the item as indicated by feedback information entries; generate sizing information for the item based on the purchase information and the feedback information; obtain first user data of a first user; determine, based on the first user data and the ML model, whether the first user should be classified as belonging to the first cluster or the second cluster; generate a recommendation message for the first user based on whether the user is to be classified as belonging to the first cluster or the second cluster according to the ML model; and provide the recommendation message to a first user device. The at least one NTCRM of claim 13 , wherein the first user data includes at least one of first body measurements associated with a user of the first user device, first demographic information associated with the user, or first preference information of the user. The at least one NTCRM of claim 14 , wherein the first user data is to be obtained via an online form of a webpage provided by a merchant service or obtained from data associated with one or more applications stored on a computer-readable medium of the first user device, wherein the one or more applications include at least one of a web browser, an application for purchasing items via the merchant service, or an application associated with a third party communication platform; or wherein at least the first body measurements are to be obtained by at least one sensor associated with the first user device. The at least one NTCRM of claim 13 , wherein: the purchase information is to include body measurement information to indicate at least one body measurement for a corresponding one of the individual users or preference information to indicate at least one preference for the corresponding ones of the individual users; and the feedback information is to include the body measurement information of the subset of the individual users, the preference information of the subset of the individual users, and return information of the subset of the individual users, wherein the return information is to indicate a reason why the item was returned to a merchant service, and wherein execution of the instructions is to cause the computing device to receive the return information from a return DB associated with the merchant service. The at least one NTCRM of claim 16 , wherein the first cluster represents items that were returned and the reason indicated by the return information is representative of a size related return, and the second cluster represents the items purchased that are not in the first cluster, and wherein a clustering operation to generate the first cluster and the second cluster is based on the body measurement information and the preference information of the individual users, and wherein, to generate the sizing information, execution of the instructions is to cause the computing device to: store, in the sizing DB, the model in association with the body measurement information and the preference information. The at least one NTCRM of claim 17 , wherein, to generate the recommendation message for the first user device, execution of the instructions is to cause the computing device to: determine, based on the first user data, a product of interest, first body measurements associated with a user of the first user device, and first preference information of the user of the first user device; obtain the model from the sizing information DB when the product of interest is determined to have one or more characteristics in common with the one or more characteristics of the item; determine first user characteristics associated with the user of the first user device based on the first preference information; determine using the model and based on at least one first user characteristic of the first user characteristics and the first body measurements, whether the first user characteristics should be classified as belonging to the first cluster or the second cluster; and generate the recommendation message to recommend the item for purchase when the first user characteristics are determined to be classified as belonging to the second cluster, or generate the recommendation message to not recommend the item for purchase when the first user characteristics are determined to be classified as belonging to the first cluster. The at least one NTCRM of claim 18 , wherein the ML model is based on a minimum body measurement value and a maximum body measurement value from the body measurement information of individual users associated with the purchase information and not within the subset of the individual users associated with the feedback information, and the execution of instructions is to cause the computing device to: use the ML model to classify the first user data as belonging to the second cluster when the first user data includes a body measurement within the minimum body measurement value and the maximum body measurement value. The at least one NTCRM of claim 18 , wherein the ML model is a support vector machine (SVM) model, and execution of the instructions is to cause the computing device to: use the purchase information and the feedback information as training data. The at least one NTCRM of claim 20 , wherein the recommendation message to not recommend the item for purchase includes a recommendation for another item to purchase based on the first user characteristics. The at least one NTCRM of claim 13 , wherein execution of the instructions is to cause the computing device to: generate the recommendation in response to receipt of a request message. The at least one NTCRM of claim 22 , wherein execution of the instructions is to cause the computing device to: receive the request message from a merchant application associated with a merchant service that is to sell the item, and wherein the request message is based on text entered into the merchant application, or the request message is based on a bar code or a quick response code obtained using at least one sensor of the first user device via the merchant application. The at least one NTCRM of claim 23 , wherein the first user data includes position information associated with the first user device, and execution of the instructions is to cause the computing device to: generate the recommendation based on a determination as to whether the distance between the item and the first user device is determined to be within a predefined distance.
CPC Classification	INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE; COMMERCIAL; FINANCIAL; MANAGERIAL OR SUPERVISORY PURPOSES;SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE; COMMERCIAL; FINANCIAL; MANAGERIAL OR SUPERVISORY PURPOSES; NOT OTHERWISE PROVIDED FOR COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
Examiner	Adam L Levine
Extended Family	012-502-083-957-225 027-976-914-313-543
Patent ID	10497043
Inventor/Author	Yarvis Mark D Uppala Anantha Deepthi
IPC	G06Q30/06 G06N5/04 G06N20/00 G06N20/10 G06Q30/00
Status	Inactive
Owner	Intel Corporation
Simple Family	012-502-083-957-225 027-976-914-313-543
CPC (with Group)	G06Q30/0631 G06N5/04 G06N20/10 G06Q30/016 G06Q30/0621 G06Q30/0623 G06N20/00
Issuing Authority	United States Patent and Trademark Office (USPTO)
Kind	Patent/New European patent specification (amended specification after opposition procedure)