CN116883181B - Financial service pushing method based on user portrait, storage medium and server - Google Patents
Financial service pushing method based on user portrait, storage medium and server Download PDFInfo
- Publication number
- CN116883181B CN116883181B CN202311147802.0A CN202311147802A CN116883181B CN 116883181 B CN116883181 B CN 116883181B CN 202311147802 A CN202311147802 A CN 202311147802A CN 116883181 B CN116883181 B CN 116883181B
- Authority
- CN
- China
- Prior art keywords
- financial
- characterization vector
- behavior
- service
- vector
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 49
- 239000013598 vector Substances 0.000 claims abstract description 566
- 238000012512 characterization method Methods 0.000 claims abstract description 513
- 238000005259 measurement Methods 0.000 claims abstract description 103
- 238000011156 evaluation Methods 0.000 claims abstract description 17
- 230000006399 behavior Effects 0.000 claims description 414
- 238000013507 mapping Methods 0.000 claims description 34
- 238000000605 extraction Methods 0.000 claims description 15
- 238000004590 computer program Methods 0.000 claims description 10
- 238000004458 analytical method Methods 0.000 claims description 9
- 238000001914 filtration Methods 0.000 claims description 7
- 238000013215 result calculation Methods 0.000 claims description 6
- 238000010801 machine learning Methods 0.000 claims description 4
- 238000005457 optimization Methods 0.000 claims description 4
- 238000004891 communication Methods 0.000 description 13
- 238000010276 construction Methods 0.000 description 12
- 239000010410 layer Substances 0.000 description 11
- 238000012216 screening Methods 0.000 description 11
- 238000012545 processing Methods 0.000 description 9
- 238000013527 convolutional neural network Methods 0.000 description 8
- 239000000284 extract Substances 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 238000013473 artificial intelligence Methods 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 230000007246 mechanism Effects 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 238000013480 data collection Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000013508 migration Methods 0.000 description 2
- 230000005012 migration Effects 0.000 description 2
- 241000288105 Grus Species 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000003542 behavioural effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000002457 bidirectional effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000010267 cellular communication Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 239000000470 constituent Substances 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 239000002356 single layer Substances 0.000 description 1
- 239000004984 smart glass Substances 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—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
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/06—Asset management; Financial planning or analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Abstract
The application provides a financial service pushing method, a storage medium and a server based on user portraits, which are characterized in that a first financial behavior characterization vector is obtained by acquiring a financial behavior log of a user to extract a financial behavior characterization vector, and a second financial behavior characterization vector is obtained by projecting the first financial behavior characterization vector; extracting a financial service characterization vector of each alternative financial service in the alternative financial service list to obtain a corresponding first financial service characterization vector, and projecting the first financial service characterization vector to obtain a corresponding second financial service characterization vector; and matching the target alternative financial business in the alternative financial business list by combining the first and second commonality evaluation results of the first and second alternative financial business to push. When the service pushing is performed, the application respectively performs commonality measurement coefficient acquisition on the financial service and the financial behavior characterization vector in the service information domain and the behavior information domain, and prevents information errors caused by matching in one value domain, thereby increasing the accuracy of the financial service pushing and improving the service experience of users.
Description
Technical Field
The application relates to the field of financial data processing, in particular to a financial service pushing method based on user portraits, a storage medium and a server.
Background
With the wide popularization of intelligent finance, the development of the internet social network is combined, and meanwhile, under the addition of new generation information technologies such as artificial intelligence, big data, cloud computing and the like, the financial industry is comprehensively promoted in business processes, business development, customer service and the like. In the business development link, financial service pushing is an important gripper for improving the conversion rate of financial products. At present, artificial intelligence is widely applied to financial product pushing in the financial field, personalized recommendation is performed by analyzing financial behaviors, preferences and demands of users and a large amount of financial data, so that the data processing efficiency is improved, the real-time performance of service pushing is also improved, and the marketing cost of financial enterprises is reduced, wherein how to accurately push financial services is always the direction of research improvement of industries, and the accuracy of current financial service pushing is still improved.
Disclosure of Invention
The application provides a financial service pushing method, a storage medium and a server based on user portraits, so as to improve the accuracy of financial service pushing.
According to an aspect of the present application, there is provided a first aspect, and an embodiment of the present application provides a financial service pushing method based on a user portrait, applied to a server, where the method includes:
acquiring a user financial behavior log corresponding to a target user, and extracting a financial behavior characterization vector from the user financial behavior log to obtain a first financial behavior characterization vector;
projecting the first financial behavior characterization vector into a financial business characterization vector field corresponding to financial business information to obtain a second financial behavior characterization vector;
extracting a financial service characterization vector of each alternative financial service in an alternative financial service list to obtain a first financial service characterization vector of each alternative financial service, wherein the alternative financial service list is matched with a user portrait of the target user;
projecting the first financial service characterization vector into a financial behavior characterization vector field corresponding to a financial behavior to obtain a second financial service characterization vector of each alternative financial service;
matching a target candidate financial service corresponding to the user financial behavior log in the candidate financial service list through a first commonality evaluation result between the first financial behavior characterization vector and the second financial behavior characterization vector and a second commonality evaluation result between the first financial behavior characterization vector and the second financial behavior characterization vector;
And pushing the target alternative financial service to terminal equipment which the target user logs in.
As an implementation manner, the obtaining a user financial behavior log, and extracting a financial behavior characterization vector from the user financial behavior log, to obtain a first financial behavior characterization vector, includes:
acquiring a user financial behavior log, and extracting a financial behavior characterization vector from the user financial behavior log based on a first gating circulation unit network after debugging is completed to obtain a plurality of intermediate financial behavior characterization vectors;
obtaining an average result calculation result of the plurality of intermediate financial behavior characterization vectors to obtain a first financial behavior characterization vector;
the projecting the first financial behavior characterization vector into a financial business characterization vector field corresponding to financial business information to obtain a second financial behavior characterization vector includes:
acquiring a financial business record characterization vector, wherein the financial business record characterization vector is a characterization vector for marking financial business semantics in financial business information when the financial business characterization vector is extracted;
and distilling the first financial behavior characterization vector to a financial business characterization vector domain through the financial business record characterization vector to obtain a second financial behavior characterization vector.
As one embodiment, the distilling the first financial behavior characterization vector to the financial behavior characterization vector domain by the financial behavior record characterization vector to obtain a second financial behavior characterization vector includes:
obtaining vector space similarity between the financial business record characterization vector and the first financial behavior characterization vector to obtain a commonality characterization vector;
classifying and mapping the commonality characterization vector through a classification and mapping network after debugging is completed to obtain a decision characterization vector;
and obtaining a second financial behavior characterization vector according to the decision characterization vector and the first financial behavior characterization vector.
As one embodiment, the obtaining a second financial behavior characterization vector according to the decision characterization vector and the first financial behavior characterization vector includes:
obtaining an average result calculation result of the decision characterization vector to obtain an average decision characterization vector;
obtaining a multiplication result of the mean decision token vector and the first financial behavior token vector to obtain a second financial behavior token vector;
the extracting the financial service characterization vector of each alternative financial service in the alternative financial service list to obtain a first financial service characterization vector of each alternative financial service includes:
Extracting financial business information characterization vectors of each alternative financial business in the alternative financial business list to obtain financial business information characterization vectors corresponding to each alternative financial business;
and carrying out characteristic information focusing analysis on the financial service information characterization vector corresponding to each alternative financial service based on the financial service record characterization vector to obtain a first financial service characterization vector of each alternative financial service.
As an implementation manner, the extracting the financial service information characterization vector of the financial service information of each candidate financial service in the candidate financial service list to obtain the financial service information characterization vector corresponding to each candidate financial service includes:
dividing each alternative financial transaction in the alternative financial transaction list into a plurality of financial transaction information;
and carrying out linear filtering on the plurality of financial business information corresponding to each alternative financial business based on a preset machine learning network to obtain a financial business information characterization vector corresponding to each alternative financial business.
As an implementation manner, the projecting the first financial service characterization vector into a financial behavior characterization vector field corresponding to a financial behavior to obtain a second financial service characterization vector of each candidate financial service includes:
Projecting the first financial service characterization vector of each alternative financial service through a debugged second gating circulation unit network to obtain a plurality of middle layer characterization vectors of each alternative financial service;
obtaining a second financial service characterization vector of each alternative financial service according to the plurality of middle layer characterization vectors of each alternative financial service and the corresponding predicted financial service behavior segment capacity;
the method comprises the steps of obtaining a financial business record characterization vector, wherein the financial business record characterization vector is before a financial business semantic in financial business information is marked when the financial business characterization vector is extracted, and the method further comprises the following steps:
acquiring a debugging and learning sample binary group, wherein the debugging and learning sample binary group comprises a financial business learning sample and a financial behavior log learning sample corresponding to the financial business learning sample;
extracting a financial service information characterization vector from the financial service learning sample, and carrying out characteristic information focusing analysis on the extracted financial service information characterization vector based on the financial service record characterization vector to be debugged to obtain a first financial service characterization vector sample of the financial service learning sample;
Loading the first financial service characterization vector sample into a first gating circulation unit network to be debugged for projection, and obtaining a second financial service characterization vector sample based on the output middle layer characterization vector;
loading the financial behavior log learning sample into a second gating circulation unit network to be debugged for projection, and obtaining an average result of the output characterization vector to obtain a first financial behavior characterization vector sample;
loading vector space similarity between the first financial behavior characterization vector sample and the financial business record characterization vector to be debugged into a classification mapping network to be debugged, and obtaining a second financial behavior characterization vector sample based on the output characterization vector of the classification mapping network to be debugged and the first financial behavior characterization vector sample;
and debugging the financial service record characterization vector to be debugged, the first gating circulation unit network, the second gating circulation unit network and the classification mapping network according to the first financial service characterization vector sample, the second financial service characterization vector sample, the first financial behavior characterization vector sample and the second financial behavior characterization vector sample.
As an implementation manner, the debugging the to-be-debugged financial service record characterization vector, the first gating circulation unit network, the second gating circulation unit network and the classification mapping network according to the first financial service characterization vector sample, the second financial service characterization vector sample, the first financial behavior characterization vector sample and the second financial behavior characterization vector sample includes:
obtaining a first learning sample commonality measurement result of the first financial service characterization vector sample and the second financial service characterization vector sample, and obtaining a second learning sample commonality measurement result of the second financial service characterization vector sample and the first financial service characterization vector sample;
obtaining an average result of the first learning sample commonality measurement result and the second learning sample commonality measurement result, and obtaining a target learning sample commonality measurement result;
acquiring a first error value through the target learning sample commonality measurement result, wherein the first error value characterizes a measurement learning error;
acquiring a second error value for supervising the second gating cycle unit network;
And carrying out back propagation on the first error value and the second error value, and carrying out optimization debugging on the to-be-debugged financial service record characterization vector, the first gating circulation unit network, the second gating circulation unit network and the classification mapping network.
As one embodiment, the matching, in the candidate financial service list, the target candidate financial service corresponding to the user financial behavior log by the first commonality evaluation result between the first financial behavior characterization vector and the second financial behavior characterization vector, and the second commonality evaluation result between the first financial behavior characterization vector and the second financial behavior characterization vector includes:
obtaining a commonality measurement coefficient between the first financial behavior characterization vector and the second financial business characterization vector to obtain a first commonality measurement coefficient;
obtaining a commonality measurement coefficient between the second financial behavior characterization vector and the first financial business characterization vector to obtain a second commonality measurement coefficient;
obtaining an average result of the first commonality measurement coefficient and the second commonality measurement coefficient to obtain a target commonality measurement coefficient of the user financial behavior log and each alternative financial business;
And determining a target alternative financial business corresponding to the user financial behavior log in the alternative financial business list according to the target commonality measurement coefficient.
According to another aspect of the present application there is provided a computer readable storage medium having stored thereon a computer program which, when run on a processor, causes the processor to perform the steps of the method as described in the first aspect above.
According to a third aspect of the present application, there is provided a server comprising: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the steps of the method of the first aspect.
The application at least has the following beneficial effects:
according to the financial service pushing method based on the user portrait, financial behavior logs of the user are obtained, financial behavior characterization vectors are extracted from the user financial behavior logs, and first financial behavior characterization vectors are obtained; projecting the first financial behavior characterization vector into a financial business characterization vector field corresponding to the financial business information to obtain a second financial behavior characterization vector; extracting a financial service characterization vector of each alternative financial service in the alternative financial service list to obtain a first financial service characterization vector of each alternative financial service; projecting the first financial service characterization vector into a financial behavior characterization vector field corresponding to the financial behavior to obtain a second financial service characterization vector of each alternative financial service; and matching a target alternative financial business corresponding to the financial behavior log of the user in the alternative financial business list through a first commonality evaluation result between the first financial behavior characterization vector and the second financial behavior characterization vector and a second commonality evaluation result between the first financial behavior characterization vector and the second financial behavior characterization vector. Based on the above, the financial service pushing method based on the user portrait, provided by the application, respectively carries out the commonality measurement coefficient acquisition on the financial service characterization vector and the financial behavior characterization vector in the service information domain and the behavior information domain when carrying out service pushing, and prevents the information error caused by matching in one value domain, thereby increasing the accuracy of financial service pushing and improving the service experience of users.
It should be understood that the description in this section is not intended to identify key or critical features of the embodiments of the application or to delineate the scope of the application. Other features of the present application will become apparent from the description that follows.
Drawings
The accompanying drawings illustrate exemplary embodiments and, together with the description, serve to explain exemplary implementations of the embodiments. The illustrated embodiments are for exemplary purposes only and do not limit the scope of the claims. Throughout the drawings, identical reference numerals designate similar, but not necessarily identical, elements.
Fig. 1 shows an application scenario schematic of a user portrayal-based financial service push method according to an embodiment of the application.
FIG. 2 illustrates a flow chart of a user portrayal-based financial service push method according to an embodiment of the application.
Fig. 3 is a schematic diagram showing a functional module architecture of a financial service pushing apparatus according to an embodiment of the present application.
Fig. 4 shows a schematic composition of a server according to an embodiment of the present application.
Detailed Description
Exemplary embodiments of the present application will now be described with reference to the accompanying drawings, in which various details of the embodiments of the present application are included to facilitate understanding, and are to be considered merely exemplary. Accordingly, one of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope of the application. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.
In the present application, the use of the terms "first," "second," etc. to describe various elements is not intended to limit the positional relationship, timing relationship, or importance relationship of these elements unless otherwise indicated, and such terms are merely used to distinguish one element from another element. In some examples, a first element and a second element may refer to the same instance of the element, and in some cases, they may also refer to different instances based on the description of the context.
The terminology used in the description of the various illustrated examples in this disclosure is for the purpose of describing particular examples only and is not intended to be limiting. Unless the context clearly indicates otherwise, the elements may be one or more if the number of the elements is not specifically limited. Furthermore, the term "and/or" as used in the present application encompasses any and all possible combinations of the listed items.
Fig. 1 shows a schematic diagram of a financial services delivery system 100 provided in accordance with an embodiment of the present application. The financial services delivery system 100 includes one or more terminal devices 101, a server 120, and one or more communication networks 110 coupling the one or more terminal devices 101 to the server 120. Terminal device 101 may be configured to execute one or more applications.
In an embodiment of the present application, server 120 may run one or more services or software applications that enable the execution of a user portrayal-based financial service push method.
In some embodiments, server 120 may also provide other services or software applications, which may include non-virtual environments and virtual environments. In some embodiments, these services may be provided as web-based services or cloud services, for example, provided to the user of terminal device 101 under a software as a service (SaaS) model.
In the configuration shown in fig. 1, server 120 may include one or more components that implement the functions performed by server 120. These components may include software components, hardware components, or a combination thereof that are executable by one or more processors. A user operating terminal device 101 may in turn utilize one or more applications to interact with server 120 to utilize the services provided by these components. It should be appreciated that a variety of different system configurations are possible, which may be different from the financial services delivery system 100. Accordingly, FIG. 1 is one example of a system for implementing the various methods described herein and is not intended to be limiting.
The user may use the terminal device 101 to generate financial actions such as page browsing of financial product information (where intermediate actions involving clicking, reading of information etc., different information, such as product pages, giving different codes), consultation, purchase, form filling, complaints etc. The terminal device 101 may provide an interface with which a user using the terminal device 101 can interact with the terminal device 101. The terminal device 101 may also output information to the user via the interface. The terminal device 101 may include various types of computer devices, such as portable handheld devices, general purpose computers (such as personal computers and laptop computers), workstation computers, wearable devices, smart screen devices, self-service terminal devices, service robots, gaming systems, thin clients, various messaging devices, and the like. These computer devices may run various types and versions of software applications and operating systems, such as MICROSOFT Windows, APPLE iOS, UNIX-like operating systems, linux, or Linux-like operating systems (e.g., GOOGLE Chrome OS); or include various mobile operating systems such as MICROSOFT Windows Mobile OS, iOS, windows Phone, android. Portable handheld devices may include cellular telephones, smart phones, tablet computers, personal Digital Assistants (PDAs), and the like. Wearable devices may include head mounted displays (such as smart glasses) and other devices. The gaming system may include various handheld gaming devices, internet-enabled gaming devices, and the like. The terminal device 101 is capable of executing various different applications, such as various Internet-related applications, communication applications (e.g., email applications), short Message Service (SMS) applications, and may use various communication protocols.
Network 110 may be any type of network known to those skilled in the art that may support data communications using any of a number of available protocols, including but not limited to TCP/IP, SNA, IPX, etc. For example only, the one or more networks 110 may be a Local Area Network (LAN), an ethernet-based network, a token ring, a Wide Area Network (WAN), the internet, a virtual network, a Virtual Private Network (VPN), an intranet, an extranet, a blockchain network, a Public Switched Telephone Network (PSTN), an infrared network, a wireless network (e.g., bluetooth, WIFI), and/or any combination of these and/or other networks.
The server 120 may include one or more general purpose computers, special purpose server computers (e.g., PC (personal computer) servers, UNIX servers, mid-end servers), blade servers, mainframe computers, server clusters, or any other suitable arrangement and/or combination. The server 120 may include one or more virtual machines running a virtual operating system, or other computing architecture that involves virtualization (e.g., one or more flexible pools of logical storage devices that may be virtualized to maintain virtual storage devices of the server). In various embodiments, server 120 may run one or more services or software applications that provide the functionality described below.
The computing units in server 120 may run one or more operating systems including any of the operating systems described above as well as any commercially available server operating systems. Server 120 may also run any of a variety of additional server applications and/or middle tier applications, including HTTP servers, FTP servers, CGI servers, JAVA servers, database servers, etc.
In some implementations, the server 120 can include one or more applications to analyze and consolidate data feeds and/or event updates received from users of the terminal devices 101. Server 120 may also include one or more applications to display data feeds and/or real-time events via one or more display devices of terminal device 101.
In some implementations, the server 120 may be a server of a distributed system or a server that incorporates a blockchain. The server 120 may also be a cloud server, or an intelligent cloud computing server or intelligent cloud host with artificial intelligence technology. The cloud server is a host product in a cloud computing service system, so as to solve the defects of large management difficulty and weak service expansibility in the traditional physical host and virtual private server (VPS, virtual Private Server) service.
The financial services push system 100 may also include one or more databases 130. In some embodiments, these databases may be used to store data and other information. For example, one or more of the databases 130 may be used to store information such as collected legal user financial behavioral data. Database 130 may reside in various locations. For example, the database used by the server 120 may be local to the server 120, or may be remote from the server 120 and may communicate with the server 120 via a network-based or dedicated connection. Database 130 may be of different types. In some embodiments, the database used by server 120 may be, for example, a relational database. One or more of these databases may store, update, and retrieve the databases and data from the databases in response to the commands. In some embodiments, one or more of databases 130 may also be used by applications to store application data. The databases used by the application may be different types of databases, such as key value stores, object stores, or conventional stores supported by the file system.
The financial services delivery system 100 of fig. 1 may be configured and operated in various ways to enable application of the various methods and apparatus described in accordance with the present application.
Referring to fig. 2, a flow chart of a method provided by an embodiment of the application may specifically include the following steps:
step S110, a user financial behavior log is obtained, and financial behavior characterization vectors are extracted from the user financial behavior log to obtain a first financial behavior characterization vector.
In the embodiment of the application, the financial behavior collected by the user in the preset behavior collection period is recorded in the user financial behavior log, for example, the financial behavior can be performed on a financial product terminal (such as a financial product application program installed on a device terminal such as a smart phone, a notebook computer, a tablet computer and the like) used by the user, for example, page browsing of financial product information (wherein, intermediate behaviors such as clicking, information reading and the like are involved, different information such as product pages are given different codes), consultation, purchase, form filling, complaint and the like are given a unique identifier and a unique timestamp, and related context information such as device information, user identification, application version and the like is recorded. It can be understood that different financial behaviors are adaptively configured based on analysis requirements, after the financial behaviors to be acquired are determined, embedded point codes are inserted into corresponding code positions according to the determined financial behaviors, and the embedded point codes can be a section of JavaScript code and can be directly embedded into webpage source codes or application programs. Next, financial activity event attributes are set, and for each buried event, additional attributes such as button name, page URL, user ID, etc. may be set to describe the financial activity event context in more detail. And finally, reporting and storing the data, wherein when a user triggers a buried point financial behavior event, the buried point code sends corresponding data to a data collection server or a third-party data analysis tool for storage and processing, so as to obtain a user financial behavior log. It should be noted that, when the technical scheme related to the application operates the buried point, relevant laws and regulations and privacy policies are followed, so that the privacy rights of the user are ensured not to be violated. In addition, excessive data collection or negative impact on user experience is avoided when selecting and managing buried financial behaviors.
After the server obtains the user financial behavior log, extracting a financial behavior characterization vector from the user financial behavior log to obtain a first financial behavior characterization vector, wherein the characterization vector characterizes a feature vector of the financial behavior. In the embodiment of the application, the extraction of the financial behavior characterization vector of the user financial behavior log can be the extraction of the characterization vector of the user financial behavior log in the space corresponding to the behavior layer.
As an implementation manner, obtaining a user financial behavior log, and extracting a financial behavior characterization vector from the user financial behavior log to obtain a first financial behavior characterization vector may include:
step A: acquiring a user financial behavior log, and extracting financial behavior characterization vectors of the user financial behavior log based on a first gating circulation unit network after debugging is completed to obtain a plurality of intermediate financial behavior characterization vectors;
and (B) step (B): and obtaining an average result calculation result of the plurality of intermediate financial behavior characterization vectors to obtain a first financial behavior characterization vector.
When the first financial behavior characterization vector of the behavior information domain is extracted from the user financial behavior log, the financial behavior characterization vector may be extracted from the user financial behavior log by a financial behavior encoder based on a gate-controlled loop unit network (GRU). The gating loop unit network in the financial behavior encoder is one or more, the GRU is a recurrent neural network for processing modeling tasks of sequence data and time sequence data, and the GRU is provided with a gating mechanism for controlling the flow of information and updating of memory. The embodiment of the application also provides a network for acquiring the behavior-business matching coefficient, wherein the financial behavior encoder based on the GRU is a module in the network for acquiring the behavior-business matching coefficient, and the GRU in the financial behavior encoder is regarded as a first gating circulation unit network for conveniently distinguishing other GRUs in the network. When the behavior-business matching coefficient acquisition network is debugged, the network internal configuration variables of the GRU in the financial behavior encoder are debugged and optimized. The tuning of the behavior-traffic matching coefficient acquisition network will be described later.
When the server acquires the user financial behavior log, the user financial behavior log and the alternative financial business are loaded into a behavior-business matching coefficient acquisition network to acquire the commonality measurement coefficient. At this time, the financial behavior encoder of the network performs the financial behavior characterization vector extraction of the behavior information domain on the acquired user financial behavior log, namely performs characterization vector extraction on the user financial behavior log based on the debugged first gating loop unit network, so as to obtain a plurality of intermediate financial behavior characterization vectors, wherein the intermediate financial behavior characterization vectors are intermediate generated financial behavior characterization vectors for transition; and obtaining average results of the plurality of intermediate financial behavior characterization vectors, namely, the results obtained by mean value calculation, so as to obtain the financial behavior characterization vector of the user financial behavior log in the behavior information domain. To distinguish from other financial behavior characterization vectors, a first financial behavior characterization vector is considered.
Step S120, projecting the first financial behavior characterization vector into a financial business characterization vector field corresponding to the financial business information to obtain a second financial behavior characterization vector.
After the financial behavior characterization vector (namely the first financial behavior characterization vector) of the user financial behavior log in the behavior information domain is extracted, the financial behavior characterization vector of the user financial behavior log in the behavior information domain and related financial business semantic features are combined, so that the financial behavior characterization vector distributed in the behavior information domain is distilled to the business information domain (or called as an intermediate domain and a hidden space), information migration is completed, and the financial behavior characterization vector in the intermediate domain is obtained. Specifically, the behavior-business matching coefficient obtaining network further comprises an intermediate domain representation vector screening network, when the financial behavior encoder extracts the financial behavior representation vector of the user financial behavior log to obtain a first financial behavior representation vector, the first financial behavior representation vector is loaded into the intermediate domain representation vector screening network to carry out representation vector projection, namely, the mapping of the features in the value domain is completed, and the financial behavior representation vector in the intermediate domain is obtained and regarded as a second financial behavior representation vector. The financial service information is an alternative financial service to be analyzed, for example, product descriptions of various financial service products, the description content can be formed by portrait labels, for example, a label A, B, C, and different labels correspond to features of one financial service product, or the description content can also be description text, which is not limited in particular.
As an implementation manner, projecting the first financial behavior characterization vector into a financial business characterization vector field corresponding to financial business information to obtain a second financial behavior characterization vector may specifically include:
step S121, a financial business record characterization vector is obtained, where the financial business record characterization vector is a characterization vector for marking the financial business semantics in the financial business information when the financial business characterization vector is extracted.
Step S122, distilling the first financial behavior characterization vector to the financial business characterization vector domain through the financial business record characterization vector to obtain a second financial behavior characterization vector.
The embodiment of the application projects the first financial behavior characterization vector of the user financial behavior log in the behavior information domain to the middle domain corresponding to the financial business, and can project the first financial behavior characterization vector based on the financial business record characterization vector. The financial business record characterization vector can be a characterization vector which is learned by the behavior-business matching coefficient acquisition network during debugging and marks financial business semantics with larger influence in financial business information during extraction of the financial business characterization vector. The characterization vector can extract key financial business semantic features in the financial behavior characterization vector in the behavior information domain to obtain a second financial behavior characterization vector of the user financial behavior log in the intermediate domain.
As one embodiment, distilling the first financial behavior characterization vector to the financial behavior characterization vector domain through the financial business record characterization vector to obtain a second financial behavior characterization vector, comprising:
step S1221, obtaining a vector space similarity between the financial business record characterization vector and the first financial behavior characterization vector, to obtain a commonality characterization vector.
Step S1222, classifying and mapping the commonality characterization vector through the classification and mapping network after debugging to obtain the decision characterization vector.
Step S1223, obtaining a second financial behavior characterization vector according to the decision characterization vector and the first financial behavior characterization vector.
In the embodiment of the present application, the intermediate domain characterization vector screening network further includes a classification mapping network, where the classification mapping network may be a fully-connected network (FC), such as an MLP, for completing classification mapping after fully-connecting the features of the previous intermediate layer. When the behavior-service matching coefficient is obtained for the network debugging, the internal configuration variables (namely various parameters of the network such as weight, bias, learning rate and the like) of the classification mapping network can be optimized and debugged at the same time, so that the classification mapping network after the debugging is obtained. The intermediate domain characterization vector selection network may comprise a classification mapping network having a single-layer structure or a multi-layer structure. When the intermediate domain characterization vector screening network acquires a first financial behavior characterization vector, cosine similarity between the financial business record characterization vector and the first financial behavior characterization vector is acquired first, and commonality measurement coefficient characteristics, namely similarity characteristics, of the financial business record characterization vector and the first financial behavior characterization vector on a channel are acquired. And obtaining a second financial behavior characterization vector through the decision characterization vector and the first financial behavior characterization vector.
As an implementation manner, obtaining the second financial behavior characterization vector according to the decision characterization vector and the first financial behavior characterization vector may specifically include:
step S12231, obtaining the average result calculation result of the decision token vector to obtain the average decision token vector.
Step S12232, obtaining the multiplication result of the mean decision token vector and the first financial behavior token vector to obtain a second financial behavior token vector.
When the second financial behavior characterization vector is obtained through the decision characterization vector and the first financial behavior characterization vector, the average result of the decision characterization vector can be obtained first to obtain the average decision characterization vector, and then the multiplication result between the average decision characterization vector and the first financial behavior characterization vector is further obtained to obtain the second financial behavior characterization vector.
Step S130, extracting the financial service characterization vector of each alternative financial service in the alternative financial service list to obtain a first financial service characterization vector of each alternative financial service, wherein the alternative financial service list is matched with the user portrait of the target user.
Under the condition that the financial behavior characterization vector of the financial behavior log of the user in the behavior information domain and the middle domain is extracted, the financial business characterization vector of the alternative financial business list can be extracted. The alternative financial services list is a services list that matches the user profile of the target user. It will be appreciated that in the embodiment of the present application, the server stores a plurality of financial service lists in advance to form a plurality of alternative financial service lists, where each financial service list corresponds to a different user portrait, and the user portrait may be generated based on a related technology, and is not limited herein, for example, a user portrait formed based on static information (such as basic information of a user), or a dynamic information user portrait generated based on historical behavior of the user, where the portrait describes, for example, investment preference, risk tolerance capability, consumption habit of the user, and the like.
The embodiment of the application can extract the first financial service characterization vector of the alternative financial service in the intermediate domain and the second financial service characterization vector in the behavior information domain.
As an implementation manner, extracting a financial service characterization vector of each candidate financial service in the candidate financial service list to obtain a first financial service characterization vector of each candidate financial service may specifically include:
step S131, extracting the financial business information characterization vector of each alternative financial business in the alternative financial business list to obtain the financial business information characterization vector corresponding to each alternative financial business.
Step S132, carrying out characteristic information focusing analysis on the financial service information characterization vector corresponding to each alternative financial service based on the financial service record characterization vector to obtain a first financial service characterization vector of each alternative financial service.
The extraction of the first financial service characterization vector of the intermediate domain is performed on the alternative financial services in the alternative financial service list, and specifically, the extraction may be performed by a financial service encoder in the acquisition network based on the behavior-service matching coefficient. For example, the server loads the alternative financial business to the financial business encoder, and performs financial business information characterization vector extraction on each piece of financial business information of the alternative financial business to obtain a plurality of financial business information characterization vectors corresponding to the alternative financial business. The financial encoder is a financial behavior encoder based on a memory mechanism, and is formed by a recording unit (a storage unit) of the financial business encoder, wherein the recording unit is used for carrying out characteristic information focusing analysis on a plurality of financial business information characterization vectors of the alternative financial business based on the learned financial business record characterization vectors when the financial business encoder is used for extracting the financial business characterization vectors of the alternative financial business, so as to obtain a first financial business characterization vector of the alternative financial business in an intermediate domain. In the feature information focusing analysis, the first financial service characterization vector may be specifically refined based on an attention mechanism.
As one embodiment, extracting a financial transaction information characterization vector of financial transaction information of each candidate financial transaction in the candidate financial transaction list to obtain a financial transaction information characterization vector corresponding to each candidate financial transaction, including:
in step S1311, each of the candidate financial transactions in the candidate financial transaction list is divided into a plurality of financial transaction information.
Step S1312, linear filtering is performed on the plurality of financial service information corresponding to each candidate financial service based on the preset machine learning network, so as to obtain a financial service information characterization vector corresponding to each candidate financial service.
In the embodiment of the application, the financial service encoder may further include a CNN, and when the financial service information representing vector of the candidate financial service is extracted, the financial service information representing vector may be obtained by performing linear filtering on the financial service information based on the CNN. The size of the filtering matrix adopted in the linear filtering is adaptively selected according to actual conditions. When the alternative financial transaction is obtained, the alternative financial transaction is partitioned into a plurality of financial transaction information (e.g., different sets of descriptive tags, each tag set corresponding to one descriptive dimension, or different descriptive text describing different transaction characteristic dimensions). And then, carrying out linear filtering on each piece of financial business information based on the CNN to obtain a financial business information characterization vector of each alternative financial business.
Step S140, projecting the first financial service characterization vector into the financial behavior characterization vector field corresponding to the financial behavior, so as to obtain a second financial service characterization vector of each candidate financial service.
After the first financial service characterization vector of the alternative financial service in the intermediate domain is extracted, the first financial service characterization vector is projected to a financial behavior characterization vector domain corresponding to the financial behavior, and a second financial service characterization vector of the alternative financial service in the financial behavior characterization vector domain is obtained.
For example, in the embodiment of the present application, the behavior-service matching coefficient obtaining network may further include a service construction network, and the first financial service characterization vector may be projected into the behavior information domain based on the service construction network to obtain the second financial service characterization vector of the alternative financial service.
As an implementation manner, projecting the first financial service characterization vector into a financial behavior characterization vector field corresponding to a financial behavior to obtain a second financial service characterization vector of each alternative financial service may specifically include:
step S141, projecting the first financial service characterization vector of each alternative financial service through the debugged second gating circulation unit network to obtain a plurality of middle layer characterization vectors of each alternative financial service.
Step S142, obtaining a second financial service characterization vector of each alternative financial service according to the plurality of middle layer characterization vectors of each alternative financial service and the corresponding predicted financial service behavior segment capacity.
In the embodiment of the application, the behavior-service matching coefficient acquisition network comprises a service construction network and can comprise a plurality of second gating circulation unit networks, and when the behavior-service matching coefficient acquisition network is debugged, the configuration variables inside the network of the plurality of second gating circulation unit networks can be optimized and debugged to obtain the debugged second gating circulation unit network. Based on the above, when the first financial service characterization vector is loaded to the service construction network, the first financial service characterization vector can be projected based on the second gating circulation unit network with multiple debugged completed, and multiple middle layer characterization vectors corresponding to each alternative financial service can be mapped. And then, carrying out mean value solution on the plurality of middle layer characterization vectors through the predicted financial business behavior segment capacity (namely the number of the included financial business behaviors) to obtain a second financial business characterization vector of each alternative financial business.
As an implementation manner, the method for obtaining the financial business record characterization vector, before the financial business record characterization vector is a characterization vector for marking the financial business semantics in the financial business information during the extraction of the financial business characterization vector, further comprises a process of network training and debugging, and specifically comprises the following steps:
Step T1, obtaining a debugging and learning sample binary group, wherein the debugging and learning sample binary group comprises a financial business learning sample and a financial behavior log learning sample corresponding to the financial business learning sample.
And step T2, extracting a financial service information characterization vector from the financial service learning sample, and carrying out characteristic information focusing analysis on the extracted financial service information characterization vector based on the financial service record characterization vector to be debugged to obtain a first financial service characterization vector sample of the financial service learning sample.
And step T3, loading the first financial service characterization vector sample into a first gating circulation unit network to be debugged for projection, and obtaining a second financial service characterization vector sample based on the output middle layer characterization vector.
And step T4, loading the financial behavior log learning sample into a second gating circulation unit network to be debugged for projection, and obtaining an average result of the output characterization vector to obtain a first financial behavior characterization vector sample.
And step T5, loading the vector space similarity between the first financial behavior characterization vector sample and the financial business record characterization vector to be debugged into a classification mapping network to be debugged, and obtaining a second financial behavior characterization vector sample based on the output characterization vector of the classification mapping network to be debugged and the first financial behavior characterization vector sample.
And step T6, debugging the financial service record characterization vector to be debugged, the first gating circulation unit network, the second gating circulation unit network and the classification mapping network according to the first financial service characterization vector sample, the second financial service characterization vector sample, the first financial behavior characterization vector sample and the second financial behavior characterization vector sample.
Based on the above, the user portrait-based financial service pushing method provided by the application is executed by the debugged financial service-behavior commonality measurement coefficient acquisition network, the financial service record characterization vector can be obtained by learning during network debugging, and the debugged first gating circulation unit network, the debugged second gating circulation unit network and the debugged classification mapping network can be obtained during the debugging of the behavior-service matching coefficient acquisition network.
The following is a description of the constituent architecture of the network and the specific procedures of debugging.
The behavior-business matching coefficient acquisition network takes a financial business information sequence and a user financial behavior log as input data, and takes a commonality measurement coefficient value of the financial business information sequence and the financial behavior log (namely, the matching degree of the financial business information sequence and the financial behavior log) as output. The network may specifically include a financial transaction encoder, a transaction-building network, a financial behavior encoder, and an intermediate domain characterization vector screening unit. The financial business encoder comprises a CNN, a recording unit and an attribute unit. The CNN is used for extracting the characterization vector of the financial business information sequence to obtain the characterization vector of the financial business information corresponding to each piece of financial business information. The CNN can be a network which is already debugged, and when the network for obtaining the behavior-service matching coefficient is debugged, the internal configuration variable of the network of the CNN is not iterated; the recording unit is used for providing a financial business record characterization vector, namely a financial business record vector, and the financial business record vector is used for marking key financial business semantic information. The attribute unit is used for acquiring focusing information of the financial service record vector and the continuous financial service information characterization vector, so that the continuous financial service information characterization vector is fused to obtain an integral characterization vector of the input financial service, namely a first financial service characterization vector; the service construction network comprises a second gating circulation unit network, for example, a plurality of second gating circulation unit networks, and the second gating circulation unit network is used for projecting the financial service characterization vectors distributed in the middle domain to the behavior information domain to obtain financial service characterization vectors in the behavior information domain, namely, second financial service characterization vectors. The financial behavior encoder comprises a first gating circulation unit network and a first operation unit, wherein the first gating circulation unit network is a bidirectional gating circulation unit network, and is used for constructing a plurality of intermediate hidden characterization vectors of a financial behavior log, and the first operation unit is used for carrying out mean value solution on the plurality of intermediate hidden characterization vectors of the financial behavior log to obtain a financial behavior characterization vector of a behavior information domain, namely a first financial behavior characterization vector. The intermediate domain characterization vector screening unit comprises a recording unit, a second operation unit, a classification mapping network and a third operation unit. The recording unit is also used for providing a financial business record characterization vector, namely a financial business record vector; the second operation unit is used for obtaining cosine similarity between the financial business record characterization vector and the first financial behavior characterization vector; the classification mapping network is used for converting cosine similarity into a decision characterization vector of the financial business record characterization vector on the channel level of the first financial behavior characterization vector; the third operation unit is used for obtaining an average result of the decision characterization vector, and obtaining a multiplication result of the average result and the first financial service characterization vector to obtain a financial behavior characterization vector of the intermediate domain, namely a second financial behavior characterization vector; the fourth operation unit is used for obtaining a first common measurement coefficient of the first financial behavior characterization vector and the second financial behavior characterization vector, obtaining a second common measurement coefficient of the second financial behavior characterization vector and the first financial behavior characterization vector, and obtaining an average result of the first common measurement coefficient and the second common measurement coefficient to obtain a common measurement coefficient of the financial behavior log and the financial behavior log.
When the neural network is debugged, firstly, a debugging learning sample (training data for training) is obtained, the debugging learning sample comprises a plurality of positive learning sample groups and negative learning sample groups, financial business and financial behavior logs in the positive learning sample groups are related financial business-financial behavior log groups, and the corresponding commonality measurement coefficient is 1, for example; the negative learning sample set is an unrelated financial business-financial behavior log set, and the corresponding commonality metric coefficient is, for example, 0. And then, based on the execution data of which the two groups of the learning sample are networks, the corresponding commonality measurement coefficient value is used as the output of the networks, the debugging of the behavior-business matching coefficient acquisition network is completed, and the behavior-business matching coefficient acquisition network after the debugging is completed is obtained. To obtain a learned financial transaction record characterization vector, a debugged first gated loop unit network, a debugged second gated loop unit network, and a debugged classification mapping network.
As an embodiment, debugging a financial service record characterization vector, a first gating circulation unit network, a second gating circulation unit network, and a classification mapping network to be debugged according to a first financial service characterization vector sample, a second financial service characterization vector sample, a first financial behavior characterization vector sample, and a second financial behavior characterization vector sample, includes:
T61, obtaining a first learning sample commonality measurement result of the first financial service characterization vector sample and the second financial behavior characterization vector sample, and obtaining a second learning sample commonality measurement result of the second financial service characterization vector sample and the first financial behavior characterization vector sample.
And T62, obtaining an average result of the first learning sample commonality measurement result and the second learning sample commonality measurement result, and obtaining a target learning sample commonality measurement result.
And T63, acquiring a first error value through the target learning sample commonality measurement result, wherein the first error value represents the measurement learning error.
And T64, acquiring a second error value for supervising the second gating cycle unit network.
And T65, carrying out counter propagation on the first error value and the second error value so as to carry out optimization debugging on the characterization vector, the first gating circulation unit network, the second gating circulation unit network and the classification mapping network of the financial service record to be debugged.
When the behavior-service matching coefficient acquisition network is debugged, the neural network is supervised and debugged through measuring the learning error and the construction error for debugging the service construction network, the distance of the active learning sample group can be reduced through measuring the learning error, the distance of the passive learning sample group can be increased, and the accuracy of the constructed behavior information domain financial service characterization vector can be increased through the construction error.
Step S150, matching the target candidate financial business corresponding to the user financial behavior log in the candidate financial business list through the first commonality evaluation result between the first financial behavior characterization vector and the second commonality evaluation result between the first financial behavior characterization vector and the second financial behavior characterization vector.
Acquiring a first financial service characterization vector of an intermediate domain and a second financial service characterization vector of a behavior information domain of each alternative financial service, and acquiring a commonality measurement coefficient of the financial service characterization vector of the intermediate domain and the financial service characterization vector of the intermediate domain in the first financial service characterization vector of the behavior information domain and the second financial service characterization vector of the behavior information domain of a user financial behavior log, namely acquiring the similarity of the first financial service characterization vector and the second financial service characterization vector, or acquiring the commonality measurement coefficient of the first financial service characterization vector and the second financial service characterization vector; meanwhile, the financial business of the behavior information domain and the financial behavior characterization vector are subjected to commonality measurement coefficient acquisition; namely, the second financial service characterization vector and the first financial behavior characterization vector are subjected to common measurement coefficient acquisition; and finally, determining a target financial service based on the common measurement coefficient of the user financial behavior log and each alternative financial service, wherein the number of the target financial services is not limited.
As one embodiment, matching a target candidate financial transaction corresponding to a user financial behavior log in a candidate financial transaction list by a first commonality evaluation result between a first financial behavior characterization vector and a second financial behavior characterization vector, and a second commonality evaluation result between the first financial behavior characterization vector and the second financial behavior characterization vector, includes:
step S151, obtaining a commonality measurement coefficient between the first financial behavior characterization vector and the second financial behavior characterization vector, and obtaining a first commonality measurement coefficient.
Step S152, obtaining a commonality measurement coefficient between the second financial behavior characterization vector and the first financial business characterization vector, so as to obtain a second commonality measurement coefficient.
Step S153, obtaining the average result of the first commonality measurement coefficient and the second commonality measurement coefficient, and obtaining the user financial behavior log and the target commonality measurement coefficient of each alternative financial business.
Step S154, determining a target candidate financial business corresponding to the user financial behavior log in the candidate financial business list according to the target commonality measurement coefficient.
After the first common measurement coefficient and the second common measurement coefficient of each alternative financial service and each user financial behavior log in different value ranges are obtained, the target common measurement coefficient of each user financial behavior log and each alternative financial service is determined based on the average result of the common measurement coefficients in the two value ranges. The transaction pushing may be performed according to the user financial behavior log and the target commonality metric coefficient for each candidate financial transaction, for example, determining the candidate financial transaction with the target commonality metric coefficient greater than the threshold value as the target financial transaction corresponding to the user financial behavior log.
Step S160, pushing the target alternative financial business to the terminal equipment logged in by the target user.
As can be seen from the summary, in the embodiment of the present application, the first financial behavior characterization vector is obtained by acquiring the user financial behavior log and extracting the financial behavior characterization vector from the user financial behavior log; projecting the first financial behavior characterization vector into a financial business characterization vector field corresponding to the financial business information to obtain a second financial behavior characterization vector; extracting a financial service characterization vector of each alternative financial service in the alternative financial service list to obtain a first financial service characterization vector of each alternative financial service; projecting the first financial service characterization vector into a financial behavior characterization vector field corresponding to the financial behavior to obtain a second financial service characterization vector of each alternative financial service; and matching a target alternative financial business corresponding to the financial behavior log of the user in the alternative financial business list through a first commonality evaluation result between the first financial behavior characterization vector and the second financial behavior characterization vector and a second commonality evaluation result between the first financial behavior characterization vector and the second financial behavior characterization vector.
Based on the above, the financial service pushing method based on the user portrait, provided by the application, can be used for respectively acquiring the commonality measurement coefficient of the financial service characterization vector and the financial behavior characterization vector in the service information domain and the behavior information domain when the service pushing is carried out, so that the information error caused by matching in one value domain is prevented, the accuracy of the financial service pushing is improved, and the service experience of the user is improved. Further, by deeply extracting the characteristic information of each of the financial business and the financial behavior log, the information of the data of one dimension in the fixed value domain is fixed, so that the information error of the financial business and the financial behavior during matching is reduced, and the recall rate of the financial business is improved.
In another embodiment of the present application, a financial service pushing method based on user portrait is provided, which is applied to a server, and specifically includes:
step S210, the server acquires a debugging learning sample doublet.
Before the financial service pushing is performed, the commonality measurement coefficient acquisition network is debugged, for example, a learning sample binary group for debugging the commonality measurement coefficient acquisition network is firstly acquired, and the learning sample binary group comprises an active learning sample binary group and a passive learning sample binary group. The positive learning sample doublet is a financial business and financial behavior log group with an association, and the negative learning sample doublet is a financial business and financial behavior log group without an association.
Step S220, the server obtains a network based on the debugging learning sample binary group debugging commonality measurement coefficient.
After obtaining the debugging and learning sample doublet, debugging the commonality measurement coefficient acquisition network through the debugging and learning sample doublet, and then pushing the debugged commonality measurement coefficient acquisition network. For example, the learning sample doublet is respectively loaded into a to-be-debugged commonality measurement coefficient acquisition network to extract a financial service characterization vector of a sample financial service in the learning sample doublet in the middle domain and a financial service characterization vector of a behavior information domain, and extract a financial behavior characterization vector of a financial behavior log learning sample in the learning sample doublet in the middle domain and a financial behavior characterization vector of the behavior information domain, and then commonality measurement coefficient data corresponding to each learning sample doublet is acquired through the characteristics. And debugging the commonality measurement coefficient acquisition network through the debugging error function to obtain a debugged commonality measurement coefficient acquisition network. In the debugging process, the first GRU in the financial behavior encoder, the second GRU in the business construction network and the network internal configuration variable of the classification mapping network in the intermediate domain characterization vector extractor are subjected to debugging optimization to obtain a debugged first GRU, a debugged second GRU and a debugged classification mapping network, and the record vectors in the record units can be learned.
In step S230, if the user financial behavior log is obtained, the server performs feature extraction on the user financial behavior log by using a financial behavior encoder in the network based on the commonality metric coefficient to obtain a first financial behavior characterization vector.
After the network for acquiring the commonality measurement coefficient is debugged, the network is called, when the user financial behavior log is acquired, the accurate commonality measurement coefficient of each financial business in the user financial behavior log and the alternative financial business list is acquired based on the neural network, and accurate business pushing is performed through the commonality measurement coefficient.
Specifically, when the user financial behavior log is obtained, the server performs feature extraction on the user financial behavior log based on a financial behavior encoder in the network to obtain a first financial behavior characterization vector of the behavior information domain. For input data, vectorization means such as single thermal coding is adopted to vectorize a financial business behavior segment, and the vectorized vector sequence is loaded to the GRU to obtain a global financial behavior characterization vector comprising the financial business behavior segment.
In step S240, the server obtains an intermediate domain representation vector screening unit in the network based on the commonality metric coefficient to screen the first financial behavior representation vector, and obtain a second financial behavior representation vector.
The first financial behavior characterization vector of the user financial behavior log in the behavior information domain can be loaded into an intermediate domain characterization vector screening unit in the commonality measurement coefficient acquisition network to finish screening so as to obtain a second financial behavior characterization vector. In the middle domain representation vector screening unit, the first financial behavior representation vector is connected with the associated financial business semantic features, representation vector screening is carried out, and the financial behavior representation vectors distributed in the behavior information domain are distilled to the middle domain to finish information migration, so that a second financial behavior representation vector located in the middle domain is obtained.
In step S250, the server obtains a financial service encoder in the network based on the commonality metric coefficient to encode each candidate financial service in the candidate financial service list, so as to obtain a first financial service characterization vector of each candidate financial service.
For each candidate financial service in the candidate financial service list, loading the candidate financial service into a financial service encoder in a commonality measurement coefficient acquisition network to perform feature extraction, namely performing intermediate domain financial service characterization vector extraction, and obtaining a first financial service characterization vector of each candidate financial service.
For example, each candidate financial transaction is divided into a plurality of financial transaction information, and then a financial transaction information characterization vector of each financial transaction information is extracted based on a preset CNN.
In step S260, the server performs behavior information domain representation vector projection on the first financial service representation vector of each candidate financial service based on the service construction network in the commonality measurement coefficient acquisition network to obtain the second financial service representation vector of each candidate financial service.
And loading the first financial service characterization vector of each alternative financial service in the intermediate domain into a financial behavior construction unit in the commonality measurement coefficient acquisition network, constructing a financial behavior mark corresponding to each alternative financial service, and projecting the financial service semantic features distributed in the intermediate domain into the behavior information domain to obtain a second financial service characterization vector positioned in the behavior information domain.
In step S270, the server obtains the commonality measurement coefficient between the first financial behavior characterization vector and the second financial business characterization vector of each candidate financial business, and obtains the first commonality measurement coefficient of the user financial behavior log and each candidate financial business.
Obtaining a first financial behavior characterization vector of a user financial behavior log in a behavior information domain and a second financial behavior characterization vector in an intermediate domain, and obtaining a commonality measurement coefficient of the user financial behavior log and each alternative financial business in each value domain after obtaining the second financial business characterization vector of each alternative financial business in the behavior information domain and the first financial business characterization vector in the intermediate domain. Specifically, the current behavior information domain is used for acquiring a user financial behavior log and a commonality measurement coefficient of each alternative financial service, namely, the commonality measurement coefficient between a first financial behavior characterization vector and a second financial service characterization vector of each alternative financial service is acquired, and the first commonality measurement coefficient of the user financial behavior log and the behavior information domain of each alternative financial service is obtained.
In step S280, the server obtains the commonality measurement coefficient between the second financial behavior characterization vector and the first financial business characterization vector of each candidate financial business, and obtains the second commonality measurement coefficient of the user financial behavior log and each candidate financial business.
And then obtaining a common measurement coefficient between the second financial behavior characterization vector of the user financial behavior log in the intermediate domain and the first financial business characterization vector of each alternative financial business in the intermediate domain, and obtaining the second common measurement coefficient of the user financial behavior log and each alternative financial business in the intermediate domain.
In step S290, the server obtains the average result of the two common measurement coefficients of the user financial behavior log and each alternative financial business, and obtains the target common measurement coefficient of the user financial behavior log and each alternative financial business.
In order to more accurately represent the common measurement coefficient of the user financial behavior log and each alternative financial service, the target common measurement coefficient of the user financial behavior log and each alternative financial service is determined based on the average result of the common measurement coefficient in different value ranges, namely, the average result of the first common measurement coefficient and the second common measurement coefficient corresponding to the user financial behavior log and each alternative financial service is obtained, and the target common measurement coefficient of the user financial behavior log and each alternative financial service is obtained.
In step S310, the server determines a target candidate financial transaction associated with the user financial transaction log according to the target commonality metric coefficient of the user financial transaction log and each candidate financial transaction.
For example, determining that at least one financial business with the maximum target commonality measurement coefficient with the user financial behavior log in the candidate financial business is the matched target candidate financial business, and pushing the target candidate financial business to the terminal equipment logged in by the target user.
According to another aspect of the present application, there is also provided a financial service pushing apparatus, referring to fig. 3, an apparatus 900 includes:
the behavior log obtaining module 910 is configured to obtain a user financial behavior log corresponding to a target user, and extract a financial behavior characterization vector from the user financial behavior log to obtain a first financial behavior characterization vector;
the first vector projection module 920 is configured to project the first financial behavior characterization vector into a financial business characterization vector field corresponding to financial business information to obtain a second financial behavior characterization vector;
a token vector extraction module 930, configured to extract a token vector of each candidate financial service in a candidate financial service list, to obtain a first token vector of each candidate financial service, where the candidate financial service list matches with a user portrait of the target user;
A second vector projection module 940, configured to project the first financial transaction token vector into a financial behavior token vector field corresponding to a financial behavior, to obtain a second financial transaction token vector of each candidate financial transaction;
a financial service matching module 950, configured to match, in the candidate financial service list, a target candidate financial service corresponding to the user financial behavior log according to a first commonality evaluation result between the first financial behavior characterization vector and the second financial behavior characterization vector and a second commonality evaluation result between the first financial behavior characterization vector and the second financial behavior characterization vector;
and the financial service pushing module 960 is configured to push the target alternative financial service to a terminal device on which the target user logs.
According to embodiments of the present application, there is also provided a server, a readable storage medium and a computer program product.
Referring to fig. 4, which is a block diagram of the structure of the server 1000 of the present application, the server 1000 includes a computing unit 1001 that can perform various appropriate actions and processes according to a computer program stored in a Read Only Memory (ROM) 1002 or a computer program loaded from a storage unit 1008 into a Random Access Memory (RAM) 1003. In the RAM 1003, various programs and data required for the operation of the server 1000 can also be stored. The computing unit 1001, the ROM 1002, and the RAM 1003 are connected to each other by a bus 1004. An input/output (I/O) interface 1005 is also connected to bus 1004.
Various components in the server 1000 are connected to the I/O interface 1005, including: an input unit 1006, an output unit 1007, a storage unit 1008, and a communication unit 10010. The input unit 1006 may be any type of device capable of inputting information to the server 1000, the input unit 1006 may receive input numeric or character information and generate key signal inputs related to user settings and/or function control of the server, and may include, but is not limited to, a mouse, a keyboard, a touch screen, a track pad, a track ball, a joystick, a microphone, and/or a remote control. The output unit 1007 may be any type of device capable of presenting information and may include, but is not limited to, a display, speakers, video/audio output terminals, vibrators, and/or printers. Storage unit 1008 may include, but is not limited to, magnetic disks, optical disks. The communication unit 10010 allows the server 1000 to exchange information/data with other devices through a computer network such as the internet and/or various telecommunications networks, and may include, but is not limited to, modems, network cards, infrared communication devices, wireless communication transceivers and/or chipsets, such as bluetooth (TM) devices, 802.11 devices, wiFi devices, wiMax devices, cellular communication devices, and/or the like.
The computing unit 1001 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of computing unit 1001 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various computing units running machine learning model algorithms, a Digital Signal Processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 1001 performs the various methods and processes described above, such as method 200. For example, in some embodiments, the method 200 may be implemented as a computer software program tangibly embodied on a machine-readable medium, such as the storage unit 1008. In some embodiments, part or all of the computer program may be loaded and/or installed onto the server 1000 via the ROM 1002 and/or the communication unit 1009. One or more of the steps of the method 200 described above may be performed when the computer program is loaded into RAM 1003 and executed by the computing unit 1001. Alternatively, in other embodiments, the computing unit 1001 may be configured to perform the method 200 in any other suitable way (e.g., by means of firmware).
Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuit systems, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), systems On Chip (SOCs), complex Programmable Logic Devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.
Program code for carrying out methods of the present application may be written in any combination of one or more programming languages. These program code may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus such that the program code, when executed by the processor or controller, causes the functions/operations specified in the flowchart and/or block diagram to be implemented. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.
In the context of the present application, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. The machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and pointing device (e.g., a mouse or trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.
The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), and the internet.
The computer system may include a client and a server. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server may be a cloud server, a server of a distributed system, or a server incorporating a blockchain.
It should be appreciated that various forms of the flows shown above may be used to reorder, add, or delete steps. For example, the steps recited in the present application may be performed in parallel, sequentially or in a different order, provided that the desired results of the disclosed embodiments are achieved, and are not limited herein.
Although embodiments or examples of the present application have been described with reference to the accompanying drawings, it is to be understood that the above-described methods, systems, and apparatus are merely illustrative embodiments or examples, and that the scope of the present application is not limited by these embodiments or examples but only by the claims following the grant and their equivalents. Various elements of the embodiments or examples may be omitted or replaced with equivalent elements thereof. Furthermore, the steps may be performed in a different order than described in the present application. Further, various elements of the embodiments or examples may be combined in various ways. It is important that as technology evolves, many of the elements described herein may be replaced by equivalent elements that appear after the application.
Claims (9)
1. A user portrayal-based financial service pushing method, applied to a server, comprising:
Acquiring a user financial behavior log corresponding to a target user, and extracting a financial behavior characterization vector from the user financial behavior log to obtain a first financial behavior characterization vector;
projecting the first financial behavior characterization vector into a financial business characterization vector field corresponding to financial business information to obtain a second financial behavior characterization vector;
extracting a financial service characterization vector of each alternative financial service in an alternative financial service list to obtain a first financial service characterization vector of each alternative financial service, wherein the alternative financial service list is matched with a user portrait of the target user;
projecting the first financial service characterization vector into a financial behavior characterization vector field corresponding to a financial behavior to obtain a second financial service characterization vector of each alternative financial service;
matching a target candidate financial service corresponding to the user financial behavior log in the candidate financial service list through a first commonality evaluation result between the first financial behavior characterization vector and the second financial behavior characterization vector and a second commonality evaluation result between the first financial behavior characterization vector and the second financial behavior characterization vector;
Pushing the target alternative financial service to terminal equipment logged in by the target user;
the method for obtaining the financial behavior log of the user comprises the steps of obtaining the financial behavior log of the user, extracting a financial behavior characterization vector of the financial behavior log of the user, obtaining a first financial behavior characterization vector, and comprising the following steps:
acquiring a user financial behavior log, and extracting a financial behavior characterization vector from the user financial behavior log based on a first gating circulation unit network after debugging is completed to obtain a plurality of intermediate financial behavior characterization vectors;
obtaining an average result calculation result of the plurality of intermediate financial behavior characterization vectors to obtain a first financial behavior characterization vector;
the projecting the first financial behavior characterization vector into a financial business characterization vector field corresponding to financial business information to obtain a second financial behavior characterization vector includes:
acquiring a financial business record characterization vector, wherein the financial business record characterization vector is a characterization vector for marking financial business semantics in financial business information when the financial business characterization vector is extracted;
distilling the first financial behavior characterization vector to a financial business characterization vector domain through the financial business record characterization vector to obtain a second financial behavior characterization vector;
The projecting the first financial service characterization vector into a financial behavior characterization vector field corresponding to a financial behavior to obtain a second financial service characterization vector of each alternative financial service, including:
projecting the first financial service characterization vector of each alternative financial service through a debugged second gating circulation unit network to obtain a plurality of middle layer characterization vectors of each alternative financial service;
and obtaining a second financial service characterization vector of each alternative financial service according to the plurality of middle layer characterization vectors of each alternative financial service and the corresponding predicted financial service behavior segment capacity.
2. The method of claim 1, wherein said distilling the first financial behavior characterization vector through the financial transaction record characterization vector to a financial transaction characterization vector domain to obtain a second financial behavior characterization vector comprises:
obtaining vector space similarity between the financial business record characterization vector and the first financial behavior characterization vector to obtain a commonality characterization vector;
classifying and mapping the commonality characterization vector through a classification and mapping network after debugging is completed to obtain a decision characterization vector;
And obtaining a second financial behavior characterization vector according to the decision characterization vector and the first financial behavior characterization vector.
3. The method of claim 2, wherein the obtaining a second financial behavior characterization vector from the decision characterization vector and the first financial behavior characterization vector comprises:
obtaining an average result calculation result of the decision characterization vector to obtain an average decision characterization vector;
obtaining a multiplication result of the mean decision token vector and the first financial behavior token vector to obtain a second financial behavior token vector;
the extracting the financial service characterization vector of each alternative financial service in the alternative financial service list to obtain a first financial service characterization vector of each alternative financial service includes:
extracting financial business information characterization vectors of each alternative financial business in the alternative financial business list to obtain financial business information characterization vectors corresponding to each alternative financial business;
and carrying out characteristic information focusing analysis on the financial service information characterization vector corresponding to each alternative financial service based on the financial service record characterization vector to obtain a first financial service characterization vector of each alternative financial service.
4. The method of claim 3, wherein the extracting the financial transaction information characterization vector for each candidate financial transaction in the candidate financial transaction list to obtain the financial transaction information characterization vector corresponding to each candidate financial transaction comprises:
dividing each alternative financial transaction in the alternative financial transaction list into a plurality of financial transaction information;
and carrying out linear filtering on the plurality of financial business information corresponding to each alternative financial business based on a preset machine learning network to obtain a financial business information characterization vector corresponding to each alternative financial business.
5. The method of claim 1, wherein the obtaining a financial transaction record characterization vector, the financial transaction record characterization vector being prior to the characterization vector marking financial transaction semantics in the financial transaction information at the time of financial transaction characterization vector extraction, further comprises:
acquiring a debugging and learning sample binary group, wherein the debugging and learning sample binary group comprises a financial business learning sample and a financial behavior log learning sample corresponding to the financial business learning sample;
extracting a financial service information characterization vector from the financial service learning sample, and carrying out characteristic information focusing analysis on the extracted financial service information characterization vector based on the financial service record characterization vector to be debugged to obtain a first financial service characterization vector sample of the financial service learning sample;
Loading the first financial service characterization vector sample into a first gating circulation unit network to be debugged for projection, and obtaining a second financial service characterization vector sample based on the output middle layer characterization vector;
loading the financial behavior log learning sample into a second gating circulation unit network to be debugged for projection, and obtaining an average result of the output characterization vector to obtain a first financial behavior characterization vector sample;
loading vector space similarity between the first financial behavior characterization vector sample and the financial business record characterization vector to be debugged into a classification mapping network to be debugged, and obtaining a second financial behavior characterization vector sample based on the output characterization vector of the classification mapping network to be debugged and the first financial behavior characterization vector sample;
and debugging the financial service record characterization vector to be debugged, the first gating circulation unit network, the second gating circulation unit network and the classification mapping network according to the first financial service characterization vector sample, the second financial service characterization vector sample, the first financial behavior characterization vector sample and the second financial behavior characterization vector sample.
6. The method of claim 5, wherein the debugging the financial transaction record characterization vector to be debugged, the first gating loop unit network, the second gating loop unit network, and the classification mapping network according to the first financial transaction characterization vector sample, the second financial transaction characterization vector sample, the first financial behavior characterization vector sample, and the second financial behavior characterization vector sample comprises:
obtaining a first learning sample commonality measurement result of the first financial service characterization vector sample and the second financial service characterization vector sample, and obtaining a second learning sample commonality measurement result of the second financial service characterization vector sample and the first financial service characterization vector sample;
obtaining an average result of the first learning sample commonality measurement result and the second learning sample commonality measurement result, and obtaining a target learning sample commonality measurement result;
acquiring a first error value through the target learning sample commonality measurement result, wherein the first error value characterizes a measurement learning error;
acquiring a second error value for supervising the second gating cycle unit network;
And carrying out back propagation on the first error value and the second error value, and carrying out optimization debugging on the to-be-debugged financial service record characterization vector, the first gating circulation unit network, the second gating circulation unit network and the classification mapping network.
7. The method of claim 1, wherein the matching of the target candidate financial transaction corresponding to the user financial transaction log in the candidate financial transaction list by the first commonality assessment result between the first financial transaction characterization vector and the second financial transaction characterization vector, and the second commonality assessment result between the first financial transaction characterization vector and the second financial transaction characterization vector, comprises:
obtaining a commonality measurement coefficient between the first financial behavior characterization vector and the second financial business characterization vector to obtain a first commonality measurement coefficient;
obtaining a commonality measurement coefficient between the second financial behavior characterization vector and the first financial business characterization vector to obtain a second commonality measurement coefficient;
obtaining an average result of the first commonality measurement coefficient and the second commonality measurement coefficient to obtain a target commonality measurement coefficient of the user financial behavior log and each alternative financial business;
And determining a target alternative financial business corresponding to the user financial behavior log in the alternative financial business list according to the target commonality measurement coefficient.
8. A computer readable storage medium, having stored thereon a computer program which, when run on a processor, causes the processor to perform the steps in the method according to any of claims 1-7.
9. A server, comprising:
at least one processor;
and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311147802.0A CN116883181B (en) | 2023-09-07 | 2023-09-07 | Financial service pushing method based on user portrait, storage medium and server |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311147802.0A CN116883181B (en) | 2023-09-07 | 2023-09-07 | Financial service pushing method based on user portrait, storage medium and server |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116883181A CN116883181A (en) | 2023-10-13 |
| CN116883181B true CN116883181B (en) | 2023-12-05 |
Family
ID=88272037
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311147802.0A Active CN116883181B (en) | 2023-09-07 | 2023-09-07 | Financial service pushing method based on user portrait, storage medium and server |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116883181B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106056407A (en) * | 2016-06-03 | 2016-10-26 | 北京网智天元科技股份有限公司 | Online banking user portrait drawing method and equipment based on user behavior analysis |
| CN108121824A (en) * | 2018-01-12 | 2018-06-05 | 北京融快线科技有限公司 | A kind of chat robots and system towards financial service |
| CN113627424A (en) * | 2021-07-14 | 2021-11-09 | 重庆师范大学 | Collaborative gating circulation fusion LSTM image labeling method |
| CN114625924A (en) * | 2022-05-16 | 2022-06-14 | 浙江大学 | Method and system for searching infringement video based on multi-vision expert knowledge distillation |
| CN115269997A (en) * | 2022-08-18 | 2022-11-01 | 中国银行股份有限公司 | Financial product recommendation method and device |
| CN116401522A (en) * | 2023-03-20 | 2023-07-07 | 阿里云计算有限公司 | Financial service dynamic recommendation method and device |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020023960A1 (en) * | 2018-07-27 | 2020-01-30 | Rocky Mountain Innovation Insights Llc | Cloud-based, data-driven artificial intelligence and machine learning financial planning and analysis visualization platform |
-
2023
- 2023-09-07 CN CN202311147802.0A patent/CN116883181B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106056407A (en) * | 2016-06-03 | 2016-10-26 | 北京网智天元科技股份有限公司 | Online banking user portrait drawing method and equipment based on user behavior analysis |
| CN108121824A (en) * | 2018-01-12 | 2018-06-05 | 北京融快线科技有限公司 | A kind of chat robots and system towards financial service |
| CN113627424A (en) * | 2021-07-14 | 2021-11-09 | 重庆师范大学 | Collaborative gating circulation fusion LSTM image labeling method |
| CN114625924A (en) * | 2022-05-16 | 2022-06-14 | 浙江大学 | Method and system for searching infringement video based on multi-vision expert knowledge distillation |
| CN115269997A (en) * | 2022-08-18 | 2022-11-01 | 中国银行股份有限公司 | Financial product recommendation method and device |
| CN116401522A (en) * | 2023-03-20 | 2023-07-07 | 阿里云计算有限公司 | Financial service dynamic recommendation method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116883181A (en) | 2023-10-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10705795B2 (en) | Duplicate and similar bug report detection and retrieval using neural networks | |
| US20200110842A1 (en) | Techniques to process search queries and perform contextual searches | |
| AU2021203164A1 (en) | Predictive issue detection | |
| CN112579909A (en) | Object recommendation method and device, computer equipment and medium | |
| CN112857268B (en) | Object area measuring method, device, electronic equipment and storage medium | |
| US10915752B2 (en) | Computer vision based asset evaluation | |
| CN114791982B (en) | Object recommendation method and device | |
| CN111125529A (en) | Product matching method and device, computer equipment and storage medium | |
| CN114443989B (en) | Ranking method, training method and device of ranking model, electronic equipment and medium | |
| US11651291B2 (en) | Real-time predictions based on machine learning models | |
| CN113868453B (en) | Object recommendation method and device | |
| CN116883181B (en) | Financial service pushing method based on user portrait, storage medium and server | |
| CN115759100A (en) | Data processing method, device, equipment and medium | |
| CN112905743B (en) | Text object detection method, device, electronic equipment and storage medium | |
| CN114445147A (en) | Electronic ticket issuing method, electronic ticket issuing device, electronic ticket issuing apparatus, and electronic ticket issuing medium | |
| CN114429678A (en) | Model training method and device, electronic device and medium | |
| CN114048383A (en) | Information recommendation method and device, electronic equipment and medium | |
| CN113312554A (en) | Method and device for evaluating recommendation system, electronic equipment and medium | |
| CN115203194A (en) | Metadata information generation method, related device, equipment and storage medium | |
| CN112579587A (en) | Data cleaning method and device, equipment and storage medium | |
| CN113963234B (en) | Data annotation processing method, device, electronic equipment and medium | |
| CN116070711B (en) | Data processing method, device, electronic equipment and storage medium | |
| CN115033782B (en) | Object recommendation method, training method, device and equipment of machine learning model | |
| CN113836939B (en) | Text-based data analysis method and device | |
| CN115809364B (en) | Object recommendation method and model training method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |