US-20260127246-A1 - BEHAVIOR-BASED USER ACCOUNTS DECOMPOSITION

US20260127246A1US 20260127246 A1US20260127246 A1US 20260127246A1US-20260127246-A1

Abstract

Methods and systems are presented for identifying different users who share a user account with an online service provider and dynamically processing transactions for the user account differently based on which user initiates the transaction request. In some embodiments, an account decomposition system may decompose the user account into distinct users who share the user account. The account decomposition system may identify different users who are sharing a user account by analyzing past transactions associated with the user account and different user devices that were used to conduct the past transactions. The account decomposition system may determine different user profiles for the different users, and may use the different user profiles to process incoming transaction requests initiated by different users of the user account.

Inventors

Tomer Handelman
Itay Margolin

Assignees

PAYPAL, INC.

Dates

Publication Date: 20260507
Application Date: 20251219

Claims (20)

1 . A system, comprising: a non-transitory memory storing instructions; and one or more hardware processors coupled with the non-transitory memory and configured to execute the instructions from the non-transitory memory to cause the system to: receive a request for processing a transaction through a user account; determine that the user account is associated with a plurality of user profiles corresponding to a plurality of users of the user account; select, from the plurality of user profiles associated with the user account, a first user profile for processing the transaction through the user account based on a plurality of clusters generated for the user account, wherein the plurality of clusters represents past transactions conducted through the user account via a plurality of user devices, wherein the past transactions are grouped into the plurality of clusters using a constrained clustering technique and based on a constraint that is associated with one or more device attributes, wherein a first subset of the past transactions having a common attribute value corresponding to the constraint is grouped in a same cluster, and wherein a second subset of the past transactions having different attribute values corresponding to the constraint is grouped in different clusters; and process the transaction using the first user profile.
2 . The system of claim 1 , wherein executing the instructions further causes the system to: map, using a machine learning model, the transaction to a vector within a multi-dimensional space based on words associated with the transaction, wherein processing the transaction is based on the vector.
3 . The system of claim 2 , wherein the machine learning model is configured to generate linguistic contexts of the words associated with the transaction and translate the linguistic contexts to the vector in the multi-dimensional space.
4 . The system of claim 2 , wherein the first user profile corresponds to a first user from the plurality of users, wherein the first user profile identifies a first cluster from the plurality of clusters that includes a portion of the past transactions conducted by the first user, wherein executing the instructions further causes the system to: determine, using the machine learning model, a benchmark vector for the first user profile based on analyzing the portion of the past transactions conducted by the first user; and compare the vector associated with the transaction against the benchmark vector determined for the first user profile.
5 . The system of claim 4 , wherein executing the instructions further causes the system to: determine a risk score for the transaction based on a difference between the vector and the benchmark vector determined for the first user profile, wherein processing the transaction is further based on the risk score.
6 . The system of claim 4 , wherein the request is received from a device, and wherein executing the instructions further causes the system to: in response to determining that a difference between the vector and the benchmark vector exceeds a threshold, prompt a user of the device for an additional credential for accessing the user account.
7 . The system of claim 1 , wherein the one or more device attributes comprise at least one of a device identifier, a screen resolution, a memory capacity, or a user interaction pattern.
8 . A method comprising: receiving, by a computer system, a transaction request associated with a user account from a first user device; determining, by the computer system, that the user account is shared among a plurality of users; determining, from the plurality of users associated with the user account, a first user who initiated the transaction request; accessing a plurality of clusters generated for the user account, wherein the plurality of clusters corresponds to a plurality of user profiles and represents past transactions conducted through the user account via a plurality of user devices, wherein the past transactions are grouped into the plurality of clusters using a constrained clustering technique based on a constraint that is associated with one or more device attributes, wherein a first subset of the past transactions having a common attribute value corresponding to the constraint are grouped in a same cluster, and wherein a second subset of the past transactions having different attribute values corresponding to the constraint are grouped in different clusters; and processing, by the computer system, the transaction request based on a first user profile from the plurality of user profiles that corresponds to the first user.
9 . The method of claim 8 , wherein the processing the transaction request comprises providing the first user device access to a computer service according to the first user profile.
10 . The method of claim 8 , wherein the processing the transaction request comprises providing customized content associated with the first user profile to the first user device.
11 . The method of claim 8 , further comprising: determining attribute values associated with the transaction request; translating the attribute values into one or more words; and determining, using a machine learning model, a deviation between the transaction request and a portion of the past transactions associated with a first cluster from the plurality of clusters identified by the first user profile based on the one or more words, wherein the processing the transaction request is further based on the deviation.
12 . The method of claim 11 , wherein the machine learning model is configured to generate linguistic contexts for the transaction request based on the one or more words and translate the linguistic contexts to the vector in a multi-dimensional space.
13 . The method of claim 12 , wherein the machine learning model is further configured to determine a risk score for the transaction request based on a deviation between the vector and a benchmark vector generated based on the portion of the past transactions.
14 . The method of claim 8 , further comprising: identifying, from the past transactions, a portion of the past transactions associated with a first cluster from the plurality of clusters corresponding to the first user profile; and determining, using a machine learning model, whether attributes associated with the transaction request deviate from a transaction pattern derived from the portion of the past transactions by more than a threshold.
15 . A non-transitory machine-readable medium having stored thereon machine-readable instructions executable to cause a machine to perform operations comprising: accessing a request for processing a transaction through a user account; determining that the user account is associated with a plurality of user profiles corresponding to a plurality of users of the user account; selecting, from the plurality of user profiles associated with the user account, a first user profile for processing the transaction through the user account based on a plurality of clusters generated for the user account, wherein the plurality of clusters comprise past transactions conducted through the user account via a plurality of user devices, wherein the past transactions are grouped into the plurality of clusters using a constrained clustering technique and based on a constraint that is associated with one or more device attributes, wherein a first subset of the past transactions having a common attribute value corresponding to the constraint are grouped in a same cluster, and wherein a second subset of the past transactions having different attribute values corresponding to the constraint are grouped in different clusters; and processing the transaction using the first user profile.
16 . The non-transitory machine-readable medium of claim 15 , wherein the first user profile identifies a first cluster from the plurality of clusters, wherein the operations further comprise: determining whether to authorize the transaction based on attribute values associated with the transaction and a portion of the past transactions associated with the first cluster.
17 . The non-transitory machine-readable medium of claim 16 , wherein the operations further comprise: determining to authorize the transaction based on the attribute values and the portion of the past transactions; and granting a user access to a computer service.
18 . The non-transitory machine-readable medium of claim 16 , wherein the operations further comprise: determining to authorize the transaction based on the attribute values and the portion of the past transactions; and providing content associated with the first user profile to a device.
19 . The non-transitory machine-readable medium of claim 16 , wherein the operations further comprise: determining not to authorize the transaction based on the attribute values and the portion of the past transactions; and prompting a user of a device for an additional credential associated with the first user profile.
20 . The non-transitory machine-readable medium of claim 16 , wherein the operations further comprise: determining to authorize the transaction based on the additional credential.

Description

CROSS REFERENCE TO RELATED APPLICATIONS This application is a continuation of U.S. patent application Ser. No. 16/938,670, filed Jul. 24, 2020, which is incorporated herein by reference in its entirety. BACKGROUND The present specification generally relates to online security, and more specifically, to dynamically providing different authentication and/or transaction processing based on different behavioral profiles associated with an account, according to various embodiments of the disclosure. RELATED ART It is common for multiple people to share a user account for accessing services and content associated with a service provider. For example, members of a family may share a payment account with a payment service provider for facilitating electronic payment transactions for the family. In another example, members of a household may share a user account with a content provider, such as Netflix®, Hulu®, etc. for accessing various content. Sharing of user accounts by multiple people can impose unique challenges to online service providers. For example, when processing transactions associated with a user account, an online service provider may use transaction behavior derived from past transactions associated with a user account to authenticate a user. Thus, when different users who exhibit different transaction behaviors share the same user account, the service provider may not be able to accurately authenticate one or more of the authorized users. Furthermore, the service provider may mistakenly authorize a fraudulent transaction request submitted by a malicious user or mistakenly deny a legitimate transaction request submitted by a legitimate user of the user account due to the inconsistent transaction behaviors exhibited by the different users of the user account. Thus, there is a need to provide a transaction processing system that can identify different users who share the same user account. BRIEF DESCRIPTION OF THE FIGURES FIG. 1 is a block diagram illustrating an electronic transaction system according to an embodiment of the present disclosure; FIG. 2 is a block diagram illustrating an account decomposition module according to an embodiment of the present disclosure; FIG. 3 illustrates clustering of past transactions according to an embodiment of the present disclosure; FIG. 4 illustrates the dynamic selection of user profiles for processing different transaction requests associated with a user account according to an embodiment of the present disclosure; FIG. 5 is a flowchart showing a process of user account decomposition according to an embodiment of the present disclosure; and FIG. 6 is a block diagram of a system for implementing a device according to an embodiment of the present disclosure. Embodiments of the present disclosure and their advantages are best understood by referring to the detailed description that follows. It should be appreciated that like reference numerals are used to identify like elements illustrated in one or more of the figures, wherein showings therein are for purposes of illustrating embodiments of the present disclosure and not for purposes of limiting the same. DETAILED DESCRIPTION The present disclosure describes methods and systems for identifying different users who share a user account with an online service provider and dynamically processing transactions for the user account differently based on which user initiates the transaction request. As discussed above, an online service provider, such as a payment service provider, an online content provider, etc., may use transaction behavior derived from past transactions of a user account to determine how to process a transaction request for the user account. For example, the online service provider may determine attributes associated with the past transactions (e.g., products/services purchased, times of purchase, purchase amounts, etc.) of the user account. The online service provider may determine transaction behavior (e.g., transaction patterns) for the user account based on the attributes associated with the past transactions. The online service provider may then use the derived transaction behavior to process incoming transaction requests associated with the user account. In some embodiments, the online service provider may determine a risk of an incoming transaction request for the user account based on whether the incoming transaction request matches the transaction patterns derived for the user account (e.g., is the product being purchased in the same product categories of products in the past transactions, etc.). The online service provider may process the incoming transaction request (e.g., authorize, deny, etc.) based on the risk. However, when multiple users conduct transactions through the same user account, the past transactions associated with the user account may exhibit erratic (e.g., inconsistent) behavior, which may cause the online service provider to determine inaccurate transaction patterns for t