WO2025011000A1

WO2025011000A1 - Method and system for providing supplemental data using chatbot

Info

Publication number: WO2025011000A1
Application number: PCT/CN2024/073580
Authority: WO
Inventors: Ang Li; Larry MA; Xie He; Tao Li; Xiaonan XIE; Kanyan YANG; Vanessa Yap; Ximei QI; Sheng Pei
Original assignee: Visa International Service Association
Priority date: 2023-07-12
Filing date: 2024-01-23
Publication date: 2025-01-16

Abstract

A method is disclosed. The method includes receiving, by a controller server computer, from a communication device of a user, a request to initiate a chat bot session. The method also includes determining a credential or derivative thereof associated with the user, and then providing the credential or derivative thereof associated with the user to a recommender system. The recommender system uses the credential or derivative thereof associated with the user as an input to a machine learning model to obtain recommended supplemental data to supply to the user in the chat bot session. The method also includes receiving, by the controller server computer, recommended supplemental data from the recommender system, and then providing real time supplemental data based on the recommended supplemental data to communication device for display in the chat bot session.

Description

METHOD AND SYSTEM FOR PROVIDING SUPPLEMENTAL DATA USING CHATBOT

CROSS-REFERENCE TO RELATED APPLICATIONS

This application is a PCT application, which claims priority to U.S. provisional application no. 63/513,202, filed on July 12, 2023, which is herein incorporated by reference in its entirety.

BACKGROUND

Current methods for providing supplemental data to a user via the user’s communication device are fragmented, and provided through a multitude of user interfaces. For example, each application on a user’s communication device can communicate with a backend application server that has its own data to present to the user. This data may not be up to date, and is typically only from a single data source.

Separately, chat bots are useful for interacting with users. However, many chat bots do not provide supplemental information to an end user based upon their context or based on the most current supplemental information available.

An improved method for presenting effective and relevant supplemental data to a user is needed. Embodiments of the invention address these and other problems.

BRIEF SUMMARY

One embodiment of the invention includes a method. The method comprises: receiving, by a controller server computer, from a communication device of a user, a request to initiate a chat bot session; determining, by the controller server computer, a credential or derivative thereof associated with the user; providing, by the controller server computer, the credential or derivative thereof associated with the user to a recommender system, wherein the recommender system uses the credential or derivative thereof associated with the user as an input to a machine learning model to obtain recommended supplemental data to supply to the user in the chat bot session; receiving, by the controller server computer, recommended supplemental data from the recommender system; and providing, by the controller server computer, real time supplemental data based on the recommended supplemental data to communication device for display in the chat bot session.

Another embodiment includes a controller server computer comprising: a processor; and a computer-readable medium comprising code, executable by the processor, for performing operations comprising: receiving from a communication device of a user, a request to initiate a chat bot session, determining a credential or derivative thereof associated with the user, providing the credential or derivative thereof associated with the user to a recommender system, wherein the recommender system uses the credential or derivative thereof associated with the user as an input to a machine learning model to obtain recommended supplemental data to supply to the user in the chat bot session, receiving the recommended supplemental data from the recommender system; and providing real time supplemental data based on the recommended supplemental data to communication device for display in the chat bot session.

Another embodiment includes a system comprising: a controller server computer comprising a processor, and a computer-readable medium comprising code, executable by the processor, for performing operations comprising receiving from a communication device of a user, a request to initiate a chat bot session, determining a credential or derivative thereof associated with the user, providing the credential or derivative thereof associated with the user to a recommender system, wherein the recommender system uses the credential or derivative thereof associated with the user as an input to a machine learning model to obtain recommended supplemental data to supply to the user in the chat bot session, receiving the recommended supplemental data from the first recommender; and providing real time supplemental data based on the recommended supplemental data to communication device for display in the chat bot session; the first recommender system.

These and other embodiments are described in further detail below.

BRIEF DESCRIPTION OF THE DRAWINGS

FIGs. 1A-1B show a block diagram of a system according to an embodiment.

FIG. 2 shows a block diagram of a conversational controller server computer according to an embodiment.

FIG. 3 shows a block diagram of a remote recommender computer according to an embodiment.

FIG. 4 shows a block diagram of a communication device according to an embodiment of the invention.

FIG. 5A illustrates an example chat bot interface for a chat bot session involving a user and a chat bot.

FIG. 5B illustrates an example chat bot interface for a chat bot session involving a user and a chat bot where offer options are shown.

DETAILED DESCRIPTION

Before discussing embodiments of the invention, some description of some terms may be helpful.

A "communication device" may comprise any suitable electronic device that may be operated by a user, which may also provide remote communication capabilities to a network. A "mobile communication device" may be an example of a "communication device" that can be easily transported. Examples of remote communication capabilities include using a mobile phone (wireless) network, wireless data network (e.g., 3G, 4G or similar networks) , Wi-Fi, Wi-Max, or any other communication medium that may provide access to a network such as the Internet or a private network. Examples of mobile communication devices include mobile phones (e.g., cellular phones) , PDAs, tablet computers, net books, laptop computers, personal music players, hand-held specialized readers, etc. Further examples of mobile communication devices include wearable devices, such as smart watches, fitness bands, ankle bracelets, rings, earrings, etc., as well as automobiles with remote communication capabilities. In some embodiments, a mobile communication device can function as a payment device (e.g., a mobile communication device can store and be able to transmit payment credentials for a transaction) .

A “user” may include an individual. In some embodiments, a user may be associated with one or more personal accounts and/or mobile devices. The user may also be referred to as a cardholder, account holder, or consumer in some embodiments.

A “resource provider” may be an entity that can provide a resource such as goods, services, information, and/or access. Examples of resource providers includes merchants, data providers, transit agencies, governmental entities, venue, and dwelling operators, etc.

A “merchant” may typically be an entity that engages in transactions and can sell goods or services, or provide access to goods or services.

An "acquirer" may typically be a business entity (e.g., a commercial bank) that has a business relationship with a particular merchant or other entity. Some entities can perform both issuer and acquirer functions. Some embodiments may encompass such single entity issuer-acquirers. An acquirer may operate an acquirer computer, which can also be generically referred to as a “transport computer. ”

An “authorizing entity” may be an entity that authorizes a request. Examples of an authorizing entity may be an issuer, a governmental agency, a document repository, an access administrator, etc.

An “issuer” may typically refer to a business entity (e.g., a bank) that maintains an account for a user. An issuer may also issue payment credentials stored on a portable device, such as a cellular telephone, smart card, tablet, or laptop to the consumer.

The term “artificial intelligence model” or “machine learning model” can include a model that may be used to predict outcomes to achieve a pre-defined goal. A machine learning model may be developed using a learning algorithm, in which training data is classified based on known or inferred patterns.

"Machine learning" can include an artificial intelligence process in which software applications may be trained to make accurate predictions through learning. The predictions can be generated by applying input data to a predictive model formed from performing statistical analyses on aggregated data. A model can be trained using training data, such that the model may be used to make accurate predictions. The prediction can be, for example, a classification of an image (e.g., identifying images of cats on the Internet) or as another example, a recommendation (e.g., a movie that a user may like or a restaurant that a consumer might enjoy) .

A “model” can include a computer program that is designed to simulate what might occur in a situation given various inputs. A model can be a machine learning model. A model can receive input data and determine an output. User features and service provider features can be input into a model to determine an output of a cart.

A “machine learning model” may include an application of artificial intelligence that provides systems with the ability to automatically learn and improve from experience without explicitly being programmed. A machine learning model may include a set of software routines and parameters that can predict an output of a process (e.g., identification of an attacker of a computer network, authentication of a computer, a suitable recommendation based on a user search query, etc. ) based on feature vectors or other input data. A structure of the software routines (e.g., number of subroutines and the relation between them) and/or the values of the parameters can be determined in a training process, which can use actual results of the process that is being modeled, e.g., the identification of different classes of input data. Examples of machine learning models include support vector machines (SVM) , models that classify data by establishing a gap or boundary between inputs of different classifications, as well as neural networks, collections of artificial “neurons” that perform functions by activating in response to inputs.

A “feature vector” may include a set of measurable properties (or “features” ) that represent some object or entity. A feature vector can include collections of data represented digitally in an array or vector structure. A feature vector can also include collections of data that can be represented as a mathematical vector, on which vector operations such as the scalar product can be performed. A feature vector can be determined or generated from input data. A feature vector can be used as the input to a machine learning model, such that the machine learning model produces some output or classification. The construction of a feature vector can be accomplished in a variety of ways, based on the nature of the input data. For example, for a machine learning classifier that classifies words as correctly spelled or incorrectly spelled, a feature vector corresponding to a word such as “LOVE” could be represented as the vector (12, 15, 22, 5) , corresponding to the alphabetical index of each letter in the input data word. For a more complex “input, ” such as a human entity, an exemplary feature vector could include features such as the human’s age, height, weight, a numerical representation of relative happiness, etc. Feature vectors can be represented and stored electronically in a feature store. Further, a feature vector can be normalized, i.e., be made to have unit magnitude. As an example, the feature vector (12, 15, 22, 5) corresponding to “LOVE” could be normalized to approximately (0.40, 0.51, 0.74, 0.17) .

“Vector representations” can include vectors which represent something. In some embodiments, vector representations can include vectors which represent nodes from graph data in a vector space. In some embodiments, vector representations can include embeddings.

A “processor” may include a device that processes something. In some embodiments, a processor can include any suitable data computation device or devices. A processor may comprise one or more microprocessors working together to accomplish a desired function. The processor may include a CPU comprising at least one high-speed data processor adequate to execute program components for executing user and/or system-generated requests. The CPU may be a microprocessor such as AMD’s Athlon, Duron and/or Opteron; IBM and/or Motorola’s PowerPC; IBM’s and Sony’s Cell processor; Intel’s Celeron, Itanium, Pentium, Xeon, and/or XScale; and/or the like processor (s) .

A “memory” may be any suitable device or devices that can store electronic data. A suitable memory may comprise a non-transitory computer readable medium that stores instructions that can be executed by a processor to implement a desired method. Examples of memories may comprise one or more memory chips, disk drives, etc. Such memories may operate using any suitable electrical, optical, and/or magnetic mode of operation.

A “server computer” may include a powerful computer or cluster of computers. For example, the server computer can be a large mainframe, a minicomputer cluster, or a group of servers functioning as a unit. In one example, the server computer may be a database server coupled to a Web server. The server computer may comprise one or more computational apparatuses and may use any of a variety of computing structures, arrangements, and compilations for servicing the requests from one or more client computers.

An “application” can include a set of computer executable instructions installed on, and executed from, a device. Applications may be installed, for example, on a communication device. An application can include any suitable functionality. For example, an application can be a communication application, an authentication application, a gaming application, an authorization application, a social media application, an information application, etc.

A “chat bot” or “chat bot platform” may be any software agent that can perform tasks or services for an individual. The chat bot may interact directly with the individual to receive input from the user (e.g., commands, in the form of speech or text) and provide output to the user (e.g., communicate, in the form of speech or text) .

Embodiments of the invention include methods and systems to provide targeted supplemental data to a user via a chat bot. The supplemental data can be in the form of offers, directions, or other supplemental information originating from multiple sources. The chat bot can behave like a concierge and present supplemental data that can be consistent with services (e.g., restaurants) that the user may wish to participate in while the user is traveling.

Some aspects of the disclosed technology provide an integrated user interface and concierge service to enable users to access multiple offers or benefits provided by a financial institution or credit card providers to a user through an integrated user interface. The user interface can include an intelligent interface to enable a user to interface or access one or more provided offers or benefits. The provided offers or benefits can be provided by an intelligent recommendation engine which can provide customized offers based on user data or user use data, such as location, prior preferences, or data about the user or related to the user obtained from another third-party application (e.g., WeChat^TM) .

Aspects of the disclosed technology allow for offers to be embedded within a natural language or other dialogue between a user and a card issuer or financial provider.

Aspects of the disclosed technology allow for the integration of offers from multiple sources into one application for the user to access. The disclosed technology can allow for the intelligent recommendation engine, which can be a machine learning or artificial intelligence-based engine, to provide the recommendations.

In some examples, aspects of the disclosed technology allow for a unified platform to incorporate all offers and benefits based on a “cross journey” or a destination country or city. The travel destination can be provided by a user through a user interface or become known through a location service enabled through a user device. The disclosed technology allows for an intelligent dialogue between a cardholder and issuer. In some examples, the intelligent dialogue can be through a chat bot with a list of actions generated using artificial intelligence or through decision trees. In other examples, the intelligent dialogue can occur through artificial intelligence or natural language generation.

In some examples, the user can choose types of content sources or offers to be integrated into the integrated user interface. In other examples, the integrated user interface can include information about coupon codes, links, car privilege (e.g., airport lounge access) , hosted purchase experiences, or card-linked offers or campaigns (e.g., discounts at national parks, museums, or coffee shops) .

FIGs. 1A and 1B show subsystems that can be parts of a system according to an embodiment. The subsystems in FIGs. 1A and 1B can also be in separate geographic regions. The subsystem in FIG. 1A can be a local subsystem in one country (e.g., China) , while the subsystem in FIG. 1B can be a remote subsystem in another country (e.g., the United States) .

FIG. 1A shows a chat bot system 110 in communication with a first recommender system 128, a second recommender system 136, a local supplemental data platform 124, a location determination system 114, and a communication device 104 operated by a user.

The location determination system 114 can be configured to provide location information (e.g., geolocation information such as latitude and longitude) in response to receiving user information and/or communication device information of the communication device 104. The location information can be provided to the communication device or the chat bot system 110. The user 102 may have previously registered a user identifier (e.g., a phone number, credential such as an account number) with their communication device 104 with the location determination system 114, so that the location determination system can determine the location of the communication device 104 upon receiving the identifier.

The chat bot system 110 can comprise a number of components including, but not limited to a conversation controller server computer 112, a supplemental data service 116, and a recommend service 120. In some embodiments, the supplemental data service 116 and the recommend service 120 may computers or software that are separate from the conversation controller server computer 112. In other embodiments, one or both of them can be software modules in the conversation controller server computer 112.

The conversation controller server computer 112 can comprise a processor and a computer readable medium comprising code, executable by a processor, to communicate with the supplemental data service 116 and the recommend service 120. The conversational controller server computer 112 can also comprise a chat bot engine. The chat bot engine can use natural language processing to simulate a human that may be chatting with the user.

The controller server computer 112 can be in communication a first recommender system 128 and a second recommender system 136 via the recommend service 120.

The first recommender system 128 can include an application service 130, a local resource provider database 132, and machine learning models 134 with embeddings. The local resource provider database 132 can comprise resource provider data such as merchant or restaurant data. If the resource provider is a restaurant, then the data stored therein can include the name of the restaurant, the type of food offered by the restaurant, a menu of the restaurant, a location of the restaurant, operating hours, etc.

The machine learning models 134 can be current versions of machine learning models created by the machine learning model generation system 160 in FIG 1B.The machine learning model generation system 160 in FIG. 1B can train one or more machine learning models using transaction database 162 from transactions conducted via the processing network computer 146, and resource provider database 164 associated with resource providers that were involved in the transactions. The first recommender system 128 and the machine learning model generation system 160 sync machine learning models via communication line 158 (such as in an offline sync process) .

The second recommender system 136 can include also include a backup resource provider database 140 and an API 138. The backup resource provider database 140 can include the same information as the local resource provider database 132.

The local resource provider database 132 in the first recommender system 128 and the backup resource provider database 140 in the second recommender system 136 are in sync as shown by the communication line 165. The chat bot system 110 will primarily access the first recommender system 128 to obtain recommended supplemental data via communication line 161 If the first recommender system 128 is offline, then the chat bot system 110 can access the second recommender system 136. The second recommender system 136 may output resource provider information to the chat bot system 110 without using the machine learning models 134. The second recommender system 136 can be used for simple queries from the user such as when the user specifically asks for a location of a resource provider or its operating hours.

In embodiments of the invention, the first recommender system 128 can receive a request for a recommendation from the recommend service 120 in the chat bot system 110. The request for the recommendation can include user information such as a credential or a derivative thereof such as token, or a hashed or masked version of the credential. After receiving the recommendation request, the first recommender system 128 may apply the user information to one or more of the machine learning models 134 to obtain a recommendation. Once the recommendation of one or more resource providers is obtained, then data about the recommended resource providers is retrieved from the local resource provider database 132. This information can be provided to the recommend service 120 and then presented to the user via the service application 106 in the communication device 104 via the chat bot managed by the controller server computer 112.

The supplemental data service 116 may have associated with it a first database of supplemental data and a second database of recommend records. In some embodiments, the supplemental data can be in the form of offers and/or offer campaigns. An offer campaign can be a process which can provide benefits to a user based upon their use of a particular user device in transactions such as payment transactions.

The local supplemental data platform 124 can include a supplemental data database. Referring to FIG. 1B, the local supplemental data platform 124 can be in communication with a remote supplemental data platform 144 via a gateway 142, which contains supplemental data (e.g., offers) based on transaction data generated using a processing network computer 146 (see FIG. 1B) . The local supplemental data platform 124 can include a receive, process and store various information such as supplemental data (e.g., local offers) by resource providers in the first region, user enrollment data, campaign data, user device data, etc. In some embodiments, the user device data can be a credential or a token. In other embodiments, the user device data can be in the form of a hashed and/or masked credential or token. This can be done to protect sensitive information while it is stored or in transit.

FIG. 1B shows the subsystem in the second region. As noted above, the subsystem can include a processing network computer 146, which generates transaction data and the remote supplemental data platform 144 that can be used to generate and store supplemental data based on the transaction data. For example, the transaction data can be credit or debit card transaction data. Transaction data associated with a certain users (e.g., certain type of spend pattern) may indicate that offers for certain types of restaurants are likely to be used by those users. Those offers can be identified by and stored in the remote supplemental data platform 144, and the provided to the local supplemental data platform 124 in FIG. 1A. The local supplemental data platform 124 can store local offers along with any offers it may have received from the remote supplemental data platform 144.

FIG. 1B additionally shows a machine learning model generation system 160. The machine learning model generation system 160 can be used to generate machine learning models based upon data in a transaction database 162 and a resource provider database 164. The information that is stored in these databases can come from the processing network computer 146. The remote supplemental data platform 144 and the gateway 142 are described above.

Methods can be described with respect to FIGs. 1A and 1B. In one embodiment, the method includes receiving, by a controller server computer, from a communication device of a user, a request to initiate a chat bot session. The method also includes determining, by the controller server computer, a credential or derivative thereof associated with the user, and then providing, by the controller server computer, the credential or derivative thereof associated with the user to a recommender system. The recommender system uses the credential or derivative thereof associated with the user as an input to a machine learning model to obtain recommended supplemental data to supply to the user in the chat bot session. The method also includes receiving, by the controller server computer, recommended supplemental data from the first recommender system comprising a machine learning model, and then providing real time supplemental data based on the recommended supplemental data to communication device for display in the chat bot session.

In step S10, the communication device 104 sends and the controller server computer 112 receives a request to initiate a chat bot session. The user 102 may be interacting with the service application 106, which may be an issuer application, a digital wallet application, a merchant application, or a service provider application. An option to initiate a chat bot session may be facilitated by the service application 106. The request may include information such as a device identifier for the communication device 104 or the user 102, and/or a credential or derivative of the credential associated with the user. In some embodiments, the credential can be a primary account number of PAN. In some embodiments, the derivative of the credential may be a hash of the credential and/or a masked credential (e.g., parts of the credential are masked by other characters or are omitted) . Another type of derivative of the credential can be a token.

Before or during the chat bot session, the controller server computer 112 can determine a credential or derivative thereof associated with the user. In some cases, the controller server computer 112 may map the device identifier (e.g., a mobile phone number) to a credential or derivative thereof stored by the controller server computer 112. In other embodiments, the credential or derivative thereof associated with the user can be sent in the request to initiate the chat bot session or as part of a query by the chat bot during the chat bot session.

In steps S18 and S20, the controller server computer 112 provides the credential or derivative thereof associated with the user to the first recommender system 128 via the recommend service 120 and the application service 130. The first recommender system 128 uses the credential or derivative thereof associated with the user as an input to a machine learning model 134 to obtain recommended supplemental data to supply to the user in the chat bot session. In some embodiments, the machine learning model 134 can identify one or more recommended resource providers based on the user’s credential or derivative thereof, and additional resource provider data can be retrieved from the local resource provider database 132. For example, the output of the machine learning model 134 can be “restaurant A” and the application service 130 can retrieve data associated with “restaurant A. ” Such data may include offers or coupons redeemable at restaurant A, its location, its hours of operation, etc. The output of the machine learning model 134 with the optional additional data may be the “recommended supplemental data. ”

In steps S22 and S24, the recommended supplemental data can be sent from the first recommender system 128 to the controller server computer 112 via the recommend service 120.

In step S26, the controller server computer 112 can send real time supplemental data including the recommended supplemental data to the service application 106 of the communication device 104, where it can be displayed to the user 102 in the chat session.

In some embodiments, to improve the performance of the recommended supplemental data provided by the machine learning model 134, information about the current context and/or preferences of the user 102 can be provided to the machine learning model along with the credential or derivative thereof. For example, as noted above, the communication device 104 and/or the controller server computer 112 can obtain the current location of the communication device 104. This information can be provided to the machine learning model 134 to provide context for where the user is presently located, so that more relevant supplemental data is provided to the user 102. In another example, the controller server computer 112 can obtain preferences or other information (e.g., campaign registration status) about the user 102 from the supplemental data service 116, which previously obtained this information from the processing network computer 146 in FIG. 1B. The preferences can be provided to the machine learning model 134 so that more relevant supplemental data is provided to the user 102.

In some embodiments, the real time supplemental data can include information that has been aggregated with the recommended supplemental data. Such information may include specific offers, coupons, or campaigns that may be stored at the supplemental data service 116. For example, in some cases, the recommended supplemental data may only include the name of a recommended resource provider, and the supplemental data service 118 may store information regarding a campaign in which the recommended resource provider participates. That information, along with the name of the recommended resource provider, can be presented to the user 102 via the communication device 104 as real time supplemental data. In some embodiments, the real time supplemental data can be the same as the recommended supplemental data.

FIG. 2 shows a block diagram of a conversational controller server computer 200 according to an embodiment. It can correspond to the conversational controller server computer 112 in FIG. 1A. The conversational controller server computer 200 includes a processor 202, and a non-transitory computer readable medium 204, a database 206, and a network interface 208 coupled to the coupled to the processor 202. The non-transitory computer readable medium 204 can comprise a chat bot engine 204A, a supplemental data processing module 204B, and a communication module 204C.

The computer readable medium may also comprise code, executable by the processor 202 to perform operations comprising: receiving from a communication device of a user, a request to initiate a chat bot session, determining a credential or derivative thereof associated with the user, providing the credential or derivative thereof associated with the user to a recommender system, wherein the recommender system uses the credential or derivative thereof associated with the user as an input to a machine learning model to obtain recommended supplemental data to supply to the user in the chat bot session, receiving the recommended supplemental data from the recommender system; and providing real time supplemental data based on the recommended supplemental data to communication device for display in the chat bot session.

The database 206 may store supplemental data, recommended supplemental data, or real time supplemental data.

The chat bot engine 204A may comprise code that causes the processor 202 to operate a chat bot.

The supplemental data processing module 204B may comprise code that is executable by the processor 202 to evaluate recommended supplemental data and to generate real time supplemental data in response to the evaluation. In some embodiments, additional data may be combined with the recommended supplemental data to form the real time supplemental data.

The communication module 204D and the processor 202 can allow the conversational controller server computer 200 to communicate with external entities.

The network interface 208 may include an interface that can allow the controller server computer 200 to communicate with external computers. Some examples of the network interface 208 may include a modem, a physical network interface (such as an Ethernet card or other Network Interface Card (NIC) ) , a virtual network interface, a communications port, a Personal Computer Memory Card International Association (PCMCIA) slot and card, or the like. The wireless protocols enabled by the network interface 208 may include Wi-Fi^TM. Data transferred via the network interface 208 may be in the form of signals which may be electrical, electromagnetic, optical, or any other signal capable of being received by the external communications interface (collectively referred to as “electronic signals” or “electronic messages” ) . These electronic messages that may comprise data or instructions may be provided between the network interface 208 and other devices via a communications path or channel. As noted above, any suitable communication path or channel may be used such as, for instance, a wire or cable, fiber optics, a telephone line, a cellular link, a radio frequency (RF) link, a WAN or LAN network, the Internet, or any other suitable medium.

FIG. 3 illustrates a mobile communication device 300 according to an embodiment. Mobile communication device 300 may include device hardware 304 coupled to a system memory 302.

Device hardware 304 may include a processor 306, a short range antenna 314, a long range antenna 316, input elements 310, a user interface 308, and output elements 312 (which may be part of the user interface 308) . Examples of input elements may include microphones, keypads, touchscreens, sensors, etc. Examples of output elements may include speakers, display screens, and tactile devices. The processor 306 can be implemented as one or more integrated circuits (e.g., one or more single core or multicore microprocessors and/or microcontrollers) , and is used to control the operation of mobile communication device 300. The processor 306 can execute a variety of programs in response to program code or computer-readable code stored in the system memory 302, and can maintain multiple concurrently executing programs or processes.

The long range antenna 316 may include one or more RF transceivers and/or connectors that can be used by mobile communication device 300 to communicate with other devices and/or to connect with external networks. The user interface 308 can include any combination of input and output elements to allow a user to interact with and invoke the functionalities of mobile communication device 300. The short range antenna 809 may be configured to communicate with external entities through a short range communication medium (e.g., using Bluetooth, Wi-Fi, infrared, NFC, etc. ) . The long range antenna 819 may be configured to communicate with a remote base station and a remote cellular or data network, over the air.

The system memory 302 can be implemented using any combination of any number of non-volatile memories (e.g., flash memory) and volatile memories (e.g., DRAM, SRAM) , or any other non-transitory storage medium, or a combination thereof media. The system memory 302 may store computer code, executable by the processor 805, for performing any of the functions described herein.

The system memory 302 may also store a service application 302A, an interaction application 302B, an authentication module 302C, credentials/tokens 302D, and an operating system 302E, The service application 302A may include an application that provides a service and may have the ability to present a chat bot to the user of the mobile communication device 300. The interaction application 302B may include code, executable by the processor 306, for forming a local connection or otherwise interacting with an external access device and/or a portable device. The authentication module 302C may comprise code, executable by the processor 306, to authenticate a user. This can be performed using user secrets (e.g., passwords) or user biometrics.

System memory 302 may also store credentials and/or tokens 302D. Credentials may also include information identifying the mobile communication device 300 and/or the user of the mobile communication device 300.

FIG. 4 shows a block diagram of a machine learning model generation system 400 according to an embodiment. It can correspond to the machine learning model generation system 160 in FIG. 1B. The machine learning model generation system 400 includes a processor 402, and a non-transitory computer readable medium 404, databases 406 comprising transaction data 406A and resource provider data 406B, and a network interface 408 coupled to the coupled to the processor 402.

The non-transitory computer readable medium 204 may comprise a machine learning generation module 404A and a communication module 204D. The computer readable medium may also comprise code, executable by the processor 402 to perform operations comprising training a machine learning model based on the transaction data 406A and the resource provider data 406B.

The machine learning generation module 408A may comprise code or software, executable by the processor 402, for embedding transaction data and the resource provider data. The machine learning generation module 408A, in conjunction with the procesor 402, can perform an embedding process (e.g., embed) the transaction data and the resource provider data in any suitable manner. The machine learning generation module 408A, in conjunction with the processor 402, can map discrete and/or categorical variables to a vector of continuous numbers. In some embodiments, the result of an embedding process (e.g., embedded transaction data and resource provider data) may be referred to as an embedding. The embedding can be a low-dimensional, learned continuous vector representation (s) . To construct representations of the transaction data and the resource provider data, the machine learning generation module 408A, in conjunction with the processor 402, can utilize an embedding neural network and a supervised task to learn the embedding (s) .

The embeddings determined by the machine learning generation module 408A, in conjunction with the processor 402, can be the parameters, or weights, of the neural network that are adjusted during training to minimize the loss on the supervised task. The machine learning generation module 408A, in conjunction with the processor 402, can determine the embedding weights (e.g., the representation of the transaction data and resource provider data as continuous vectors) .

The machine learning generation module 408A may comprise code or software, executable by the procesor 402, for training machine learning models (e.g., neural network models) . In some embodiments, the machine learning generation module 408A contain code that defines a machine learning model, as well as code that can enable the processor 402 to train the machine learning model. The trained machine learning model can accept feature inputs and determine an output (e.g., a classification, prediction, etc. ) for each input vector. The machine learning generation module 408A, in conjunction with the procesor 402, may use suitable machine learning models based on algorithms including, but not limited to: neural networks, decision trees, support vector methods, and K-means algorithms.

For example, the machine learning generation module 408A, in conjunction with the procesor 402, can build a mathematical model based on sample data, known as "training data" , to make predictions or decisions without being explicitly programmed to perform the task. In some embodiments, machine learning generation module 408A, in conjunction with the processor 402, can train a neural network. A neural network can be a model based on a collection of connected units or nodes called artificial neurons. Each connection (e.g., edge) can transmit information (e.g., a signal) from node to another. A node that receives a signal can process it and then signal additional nodes connected to it. In some embodiments, the signal at a connection between nodes can include a real number, and the output of each node can be computed by some non-linear function of the sum of its inputs. Nodes and edges can have a weight that adjusts as learning proceeds. The weight may increase or decrease the strength of the signal at an edge. In some embodiments, nodes may have a threshold such that the signal is only sent if the aggregate signal crosses that threshold. Different layers of the neural network may perform different kinds of transformations on their inputs. Further, signals can travel from the first layer (e.g., the input layer) , to the last layer (e.g., the output layer) , possibly after traversing middle layer (s) (e.g., hidden layer (s) ) . In some embodiments, the machine learning generation module 408A, in conjunction with the procesor 402, can train a neural network.

FIG. 5A illustrates an example chat bot interface for a chat bot session involving a user and a chat bot. In the chat bot session, the user can converse with the chat bot. As shown by chat bot information 502, the chat bot may provide information about a suggested itinerary after the user provides basic information about their desired travel plan. FIG. 5A shows three buttons including a first button 504 that will allow a user to obtain relevant offers, a second button 506 that will allow a user to “apply cards, ” and a third button 508 to see available campaigns. Selecting the “apply cards” button can be used accept an offer by providing a credential such as an account number to the chat bot engine.

FIG. 5B illustrated an example chat bot interface for a chat bot session involving a user and a chat bot where offer options are shown (e.g., after the user selects the “offers” button. The offers were generated in part by identification of various resource providers by the previously described machine learning model. The offers may also be generated based on the context of the conversation in the chat bot session.

Embodiments of the invention have several technical advantages. Embodiments of the invention can generate accurate real time recommended supplemental data for a user in real time in a chat bot session. This makes it easier for a chat bot to provide more relevant supplemental data to the user that the user is more likely to accept or interact with. Further, some embodiments, preserve privacy by synching machine learning models across different regions (such as in an offline sync process) , so that sensitive data does not have to be transmitted outside of a particular region. By transmitting different machine learning models to different regions from a machine learning model generation system, the latency of providing recommended supplemental data to a chat bot engine is decreased relative to the situation where all the chat bots in different regions need to communicate with a central machine learning system in a separate region. This is desirable in the chat bot context, since responses from the chat bot that take too long will make the chat bot seem less natural and less authentic to the user.

Any of the software components or functions described in this application, may be implemented as software code to be executed by a processor using any suitable computer language such as, for example, Java, C++, or Perl using, for example, conventional or object-oriented techniques. The software code may be stored as a series of instructions, or commands on a computer readable medium, such as a random access memory (RAM) , a read only memory (ROM) , a magnetic medium such as a hard-drive or a floppy disk, or an optical medium such as a CD-ROM. Any such computer readable medium may reside on or within a single computational apparatus and may be present on or within different computational apparatuses within a system or network.

The above description is illustrative and is not restrictive. Many variations of the invention may become apparent to those skilled in the art upon review of the disclosure. The scope of the invention can, therefore, be determined not with reference to the above description, but instead can be determined with reference to the pending claims along with their full scope or equivalents.

One or more features from any embodiment may be combined with one or more features of any other embodiment without departing from the scope of the invention.

A recitation of "a" , "an" or "the" is intended to mean "one or more" unless specifically indicated to the contrary.

All patents, patent applications, publications, and descriptions mentioned above are herein incorporated by reference in their entirety for all purposes. None is admitted to be prior art.

Claims

A method comprising:

receiving, by a controller server computer, from a communication device of a user, a request to initiate a chat bot session;

determining, by the controller server computer, a credential or derivative thereof associated with the user;

providing, by the controller server computer, the credential or derivative thereof associated with the user to a recommender system, wherein the recommender system uses the credential or derivative thereof associated with the user as an input to a machine learning model to obtain recommended supplemental data to supply to the user in the chat bot session;

receiving, by the controller server computer, the recommended supplemental data from the recommender system; and

providing, by the controller server computer, real time supplemental data based on the recommended supplemental data to communication device for display in the chat bot session.
The method of claim 1, wherein the controller server computer comprises a chat bot engine that is programmed to facilitate the chat bot session.
The method of claim 1, wherein the communication device is a mobile phone.
The method of claim 1, wherein the machine learning model comprises a neural network.
The method of claim 1, wherein providing, by the controller server computer, the credential or derivative thereof associated with the user to the recommender system, wherein the recommender system uses the credential or derivative thereof associated with the user as the input to the machine learning model to obtain the recommended supplemental data to supply to the user in the chat bot session comprises:

providing, by the controller server computer, the derivative of the credential to the recommender system, wherein the recommender system uses the derivative of the credential as the input to the machine learning model to obtain the recommended supplemental data to supply to the user in the chat bot session.
The method of claim 5, wherein the derivative of the credential is a hashed and/or masked version of the credential.
The method of claim 1, wherein the recommender system comprising the machine learning model is in a first geographic region, receives the machine learning model from a machine learning model generation system in a second geographic region, wherein the machine learning model generation system provides the machine learning model and other machine learning models to the recommender system in an offline sync process.
The method of claim 1, further comprising:

obtaining, by the controller server computer, a geolocation for the communication device; and

providing, by the controller server computer, the geolocation of the communication device to the recommender system with the credential or derivative thereof, and wherein the geolocation and the credential or derivative thereof are provided as inputs to the machine learning model to obtain the recommended supplemental data.
The method of claim 1, further comprising:

obtaining, by the controller server computer, a geolocation for the communication device; and

obtaining, by the controller server computer, additional supplemental data from a supplemental data service,

wherein the real time supplemental data is based upon the additional supplemental data and the recommended supplemental data.
The method of claim 1, wherein the machine learning model is trained on transaction data based on transactions conducted by multiple users with multiple resource providers, and resource provider data associated with the resource providers.
The method of claim 1, wherein the chat bot session involves a discussion regarding an itinerary of the user.
The method of claim 1, wherein the recommended supplemental data comprises resource providers that the user may interact with.
The method of claim 1, further comprising:

obtaining, by the controller server computer, additional supplemental data from a supplemental data service; and

wherein the real time supplemental data is based upon the additional supplemental data and the recommended supplemental data.
The method of claim 13, wherein the additional supplemental data comprises campaign data associated with a campaign process.
A controller server computer comprising:

a processor; and

a computer-readable medium comprising code, executable by the processor, for performing operations comprising:

receiving from a communication device of a user, a request to initiate a chat bot session,

determining a credential or derivative thereof associated with the user,

providing the credential or derivative thereof associated with the user to a recommender system, wherein the recommender system uses the credential or derivative thereof associated with the user as an input to a machine learning model to obtain recommended supplemental data to supply to the user in the chat bot session,

receiving the recommended supplemental data from the recommender system; and

providing real time supplemental data based on the recommended supplemental data to communication device for display in the chat bot session.
The controller server computer of claim 15, wherein providing, by the controller server computer, the credential or derivative thereof associated with the user to the recommender system, wherein the recommender system uses the credential or derivative thereof associated with the user as the input to the machine learning model to obtain the recommended supplemental data to supply to the user in the chat bot session comprises:

providing, by the controller server computer, the derivative of the credential to the recommender system, wherein the recommender system uses the derivative of the credential as the input to the machine learning model to obtain the recommended supplemental data to supply to the user in the chat bot session.
A system comprising:

a controller server computer comprising a processor, and a computer-readable medium comprising code, executable by the processor, for performing operations comprising receiving from a communication device of a user, a request to initiate a chat bot session, determining a credential or derivative thereof associated with the user, providing the credential or derivative thereof associated with the user to a recommender system, wherein the recommender system uses the credential or derivative thereof associated with the user as an input to a machine learning model to obtain recommended supplemental data to supply to the user in the chat bot session, receiving the recommended supplemental data from the recommender system; and providing real time supplemental data based on the recommended supplemental data to communication device for display in the chat bot session; and

the recommender system.
The system of claim 17, wherein the operations further comprise:

obtaining a geolocation for the communication device; and

providing the geolocation of the communication device to the recommender system with the credential or derivative thereof, and wherein the geolocation and the credential or derivative thereof are provided as inputs to the machine learning model to obtain the recommended supplemental data.
The system of claim 17, further comprising:

the communication device.
The system of claim 17, wherein the operations further comprise:

obtaining, by the controller server computer, a geolocation for the communication device; and

obtaining, by the controller server computer, additional supplemental data from a supplemental data service, wherein the real time supplemental data is based upon the additional supplemental data and the recommended supplemental data.