A method, device, equipment, medium and product for guiding financial services business

Abstract

The present invention discloses a method, device, equipment, medium, and product for guiding financial services. The method includes: using a financial services large language model to analyze business questions input by users to obtain target business items to be handled; determining the basic selection probability of each candidate machine based on the target business item, the user's historical behavior data, and the attribute information of each candidate machine; correcting the basic selection probability based on the current business handling status of each candidate machine, the location of the candidate machine, and the current location of the user to obtain the target selection probability of each candidate machine, and sorting the candidate machines according to the target selection probability so that the user can select the target machine from the candidate machines; and guiding the user to the target machine to handle the target business item. Embodiments of the present invention can improve the efficiency and experience of users in handling financial services.

Description

Technical Field

[0001] The present invention relates to the fields of artificial intelligence technology and financial technology, and in particular to a method, device, equipment, medium and product for guiding financial services business. Background Art

[0002] Nowadays, many banking services have been moved online, and users can complete some of the business transactions online. However, due to the wide variety of banking services, a wide range of users, and different authority requirements, offline outlets are still an indispensable way for users to conduct banking business. Summary of the Invention

[0003] The present invention provides a method, device, equipment, medium and product for guiding financial service business, so as to solve the problem of difficulty in selecting appropriate business processing outlets for users, thereby saving the time cost of users in processing business.

[0004] According to one aspect of the present invention, a method for guiding a financial services business is provided, comprising:

[0005] Use the financial services language model to analyze the business questions input by users and obtain the target business items to be handled;

[0006] Determining a basic selection probability for each candidate device based on the target service item, the user's historical behavior data, and attribute information of each candidate device;

[0007] The basic selection probability is modified according to the current business processing status of each candidate machine, the location of the candidate machine, and the current location of the user to obtain a target selection probability for each candidate machine, and the candidate machines are sorted according to the target selection probability so that the user can select a target machine from the candidate machines;

[0008] Guide the user to the target machine to handle the target business item.

[0009] According to another aspect of the present invention, a device for guiding a financial services service is provided, comprising:

[0010] The target business item module is used to analyze the business questions input by users using the financial services large language model to obtain the target business items to be handled;

[0011] A basic selection probability module is used to determine the basic selection probability of each candidate device based on the target service item, the user's historical behavior data and the attribute information of each candidate device;

[0012] a target selection probability module, configured to modify the basic selection probability based on the current business handling status of each candidate machine, the location of the candidate machine, and the current location of the user, to obtain a target selection probability for each candidate machine, and to sort the candidate machines according to the target selection probability so that the user can select a target machine from among the candidate machines;

[0013] The guidance module is used to guide the user to go to the target machine to handle the target business item.

[0014] According to another aspect of the present invention, a computer program product is provided, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements the method for guiding a financial services service according to any embodiment of the present invention.

[0015] According to another aspect of the present invention, an electronic device is provided, comprising:

[0016] At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the method for bootstrapping a financial services business as described in any embodiment of the present invention.

[0017] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for guiding the financial service business according to any embodiment of the present invention when executed.

[0018] According to another aspect of the present invention, a computer program product is provided, comprising a computer program / instruction, which, when executed by a processor, implements the method for guiding a financial services service according to any embodiment of the present invention.

[0019] This embodiment of the present invention analyzes user-entered business questions using a financial services language model and identifies the target financial services business items within the banking system that the user is seeking to process. This clarifies the user's business needs and facilitates accurate device recommendations based on these clear needs, improving the efficiency and accuracy of identifying the user's pending services. A basic selection probability for each candidate device is determined based on the target service item, the user's historical behavior data, and the attribute information of each candidate device. This basic selection probability is then modified based on the current service status of each candidate device, its location, and the user's current location to obtain a target selection probability for each candidate device. The candidate devices are then ranked according to the target selection probabilities, allowing the user to select a target device from the candidate devices. This reduces the time wasted and inconvenience caused by selecting inappropriate devices, improving the efficiency and satisfaction of users in handling financial services. The target service item, user historical behavior data, candidate device attribute information, and real-time service status and location information are referenced when selecting the target device, further improving the accuracy of target device selection. The comprehensive intelligent navigation system is built by combining digital twin technology and dynamic real-scene technology to provide point-to-point navigation guidance throughout the entire process from the user's current location to handling business, thereby improving the convenience and experience of users in handling financial business.

[0020] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present invention, nor is it intended to limit the scope of the present invention. Other features of the present invention will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0022] Figure 1 This is a first flow chart of a method for guiding a financial services business provided by an embodiment of the present invention;

[0023] Figure 2 This is a second flow chart of a method for guiding a financial services business provided by an embodiment of the present invention;

[0024] Figure 3 This is a schematic structural diagram of a device for guiding financial services provided by an embodiment of the present invention;

[0025] Figure 4 It is a schematic structural diagram of an electronic device implementing an embodiment of the present invention. DETAILED DESCRIPTION

[0026] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0027] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way are interchangeable where appropriate so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices. The relevant information involved in the present invention (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for display, data for analysis, etc.) are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data comply with the relevant laws, regulations and standards of the relevant region.

[0028] Figure 1 This is the first flow chart of a method for guiding financial services provided by an embodiment of the present invention. This embodiment can be applied to clarifying the user's business needs, selecting appropriate business processing outlets according to the user's business needs, guiding the user to the business processing outlets, and saving the user's time cost before processing the business. This method can be executed by a guidance device for financial services. The device can be implemented in the form of hardware and / or software. The device can be configured in an electronic device with corresponding data processing capabilities. Figure 1 As shown, the method includes:

[0029] S110: Use the financial services big language model to analyze the business questions input by the user and obtain target business items to be processed.

[0030] When users need to handle financial business, it is often difficult to know exactly what specific business items they need to handle. The financial service big language model analyzes the business problems input by the user and obtains the target business items to be handled. Specifically, the user logs in to the digital twin financial service platform and enters the user name and password. The digital twin financial service platform authenticates the user and optionally performs two-factor authentication (such as SMS verification code verification plus biometric authentication) to ensure account security. After successful identity authentication, the platform allocates corresponding resources to the user based on the user's permissions, including the user's accessible data sets, available tools and services, and the current user's personalized resource configuration. The resource allocation process is completed by the permission management system in the digital twin financial service platform to ensure that each user can only access resources within their permission range, thereby ensuring the security and efficiency of the platform.

[0031] On the digital twin financial services platform, users enter business questions based on ambiguous financial service needs. The digital twin financial services platform uses the financial services big language model to conduct in-depth analysis of these ambiguous user-entered financial service needs and identify specialized financial service business items within the banking system as target business items. Specifically, the digital twin financial services platform uses natural language processing technology to identify keywords and intent in user-entered business questions and, based on the context, conducts comprehensive semantic analysis to infer the ambiguous financial service business items the user needs to handle. The financial services big language model then engages in conversation with the user to further clarify and infer the user's specific needs. The financial services big language model uses natural language processing technology to parse user input, including syntax analysis and named entity recognition. Context management technology enables the financial services big language model to analyze and record conversation content to ensure consistency. Conversation state tracking technology records and updates conversation state, assisting the financial services big language model in understanding the evolving needs. Through reinforcement learning and supervised learning, the financial services big language model continuously optimizes conversation strategies to ensure accurate and efficient financial services, thereby inferring the precise target business items that the user actually needs to handle.

[0032] The Financial Services Big Language Model, by learning from massive amounts of financial data and historical user behavior, accurately understands natural language input from users and converts it into actionable business requirements. By analyzing the business questions entered by users, the Big Language Model identifies the specialized financial services business items within the banking system that the user is seeking to handle as target business items. This clarifies the user's business needs and facilitates accurate device recommendations based on these clear business requirements.

[0033] S120 : Determine a basic selection probability of each candidate device based on the target service item, the user's historical behavior data, and the attribute information of each candidate device.

[0034] All machines within a certain range of the user's current location are considered candidate machines. These include both self-service machines at self-service outlets and machines at business outlets, such as ATMs (Automated Teller Machines), automated teller machines (ATMs), teller counters, and intelligent teller machines (ITMs). Each candidate machine can handle banking services, but different candidates have different permissions and can handle different services. For example, ATMs and ATMs do not have permission to open accounts; machines at different business outlets also have different permissions.

[0035] Based on the target business item, the user's historical behavior data, and the attribute information of each candidate device, the compatibility of each candidate device with the target business item is determined, and the compatibility of each candidate device with the target business item is used as the basic selection probability of the candidate device. The basic selection probability is used to indicate whether the candidate device can handle the target business item. Among them, the user's historical behavior data refers to multi-dimensional data such as the user's historical operation records, historical business processing records, and historical feedback and evaluation in the digital twin financial service platform. The attribute information of the candidate device includes the professional financial service business items within the banking system that the candidate device can execute and the candidate device's permissions.

[0036] S130. Modify the basic selection probability according to the current business processing status of each candidate machine, the location of the candidate machine, and the current location of the user to obtain a target selection probability for each candidate machine, and sort the candidate machines according to the target selection probability so that the user can select a target machine from the candidate machines.

[0037] The distance between the location of the candidate machine and the user's current location, as well as the current business handling status of the candidate machine, will affect the processing efficiency of the target business item. For example, if the distance between the location of the candidate machine and the user's current location is very far, the user will need to waste a lot of time to reach the candidate machine, or the candidate machine is currently busy with business handling, and the user will need to wait for a long time before being able to handle the target business item when arriving at the candidate machine, which will lead to low processing efficiency of the target business item. According to the current business handling status of each candidate machine, the location of the candidate machine and the current location of the user, the basic selection probability of the candidate machine is corrected to obtain the target selection probability of each candidate machine, and the candidate machines are sorted according to the target selection probability of each candidate machine, and the sorted candidate machine list is displayed to the user for selection. The candidate machine selected by the user is used as the target machine.

[0038] S140: Guide the user to the target machine to handle the target business item.

[0039] Using dynamic real-world technology, users are intelligently guided to their target facility to complete their desired transaction. Specifically, a camera is activated to capture the user's surroundings in real time, and navigation information is overlaid on top of the real-world scene using AR (augmented reality) technology. Based on the user's current location and the location of the target facility, an optimal route is generated and displayed to the user using virtual arrows, lines, or other visual elements within the real-world scene. During navigation, the route is dynamically adjusted based on the user's actual movement trajectory and environmental changes (such as obstacles and changes in crowd density) to ensure accurate and efficient guidance. In addition to visual guidance, voice navigation is also provided to help users easily follow the instructions even in complex environments. Guiding users to their target facility using AR technology helps users quickly and accurately find the facility and conduct their desired transaction. By combining digital twin technology with dynamic real-world technology, an intelligent navigation system is constructed that provides end-to-end navigation guidance from the user's current location to the transaction, significantly improving the convenience and user experience of handling financial transactions.

[0040] In the prior art, when a user decides to go to a bank branch to handle a business, they usually choose the nearest corresponding bank branch they can find to handle the business. However, this approach often has the following disadvantages:

[0041] If a user wants to handle a business item, they can do so at a 24-hour self-service outlet at a self-service machine (such as an ATM (Automated Teller Machine), an automatic teller machine, etc.), without having to go to a manual business outlet at a machine (manual counter, intelligent teller machine, etc.). However, this information is not known to the user before arriving at a manual business outlet.

[0042] After the user arrives at a manual business outlet, he or she finds that this outlet does not have the authority to handle the business item the user wants to handle, so the user needs to spend time going to another outlet again.

[0043] After entering a manual business outlet, users need to consult repeatedly to find out whether they need to queue up at a manual counter or at an intelligent teller machine (ATM) to handle the business they want to do. In addition, intelligent teller machines are divided into different types, which need to be distinguished before handling business at the intelligent teller machine, increasing the time cost for users before handling business.

[0044] This embodiment of the present invention analyzes user-entered business questions using a financial services language model and identifies the target financial services business items within the banking system that the user is seeking to process. This clarifies the user's business needs and facilitates accurate device recommendations based on these clear needs, improving the efficiency and accuracy of identifying the user's pending services. A basic selection probability for each candidate device is determined based on the target service item, the user's historical behavior data, and the attribute information of each candidate device. This basic selection probability is then modified based on the current service status of each candidate device, its location, and the user's current location to obtain a target selection probability for each candidate device. The candidate devices are then ranked according to the target selection probabilities, allowing the user to select a target device from the candidate devices. This reduces the time wasted and inconvenience caused by selecting inappropriate devices, improving the efficiency and satisfaction of users in handling financial services. The target service item, user historical behavior data, candidate device attribute information, and real-time service status and location information are referenced when selecting the target device, further improving the accuracy of target device selection. The comprehensive intelligent navigation system is built by combining digital twin technology and dynamic real-scene technology to provide point-to-point navigation guidance throughout the entire process from the user's current location to handling business, thereby improving the convenience and experience of users in handling financial business.

[0045] In an optional embodiment, the use of the financial services large language model to analyze the business problem input by the user to obtain the target business item to be processed includes: using the financial services large language model to process the business problem input by the user to obtain the fuzzy business item to be processed, and obtaining the target sub-model associated with the fuzzy business item from each candidate sub-model in the financial services large language model; calling the target sub-model to interact with the user to obtain the target business item to be processed.

[0046] Specifically, the training process for the financial services language model requires a dataset specifically tailored to the banking industry, including a large collection of banking terminology, frequently asked questions, and user conversation logs. This dataset must cover a variety of banking scenarios, including account opening, transfers, loans, and credit card applications. The financial services language model is first pre-trained on a general language model and then fine-tuned on a banking data set. This ensures that the model possesses both high versatility and specificity, enabling it to better understand and process banking-related natural language input.

[0047] As banking services expand, banking datasets also become richer. To maintain the accuracy of the financial services language model, regular incremental training of the model is required. Using a modular component design, the financial services language model is divided into modular components. During incremental learning, only the parameters of specific sub-modules need to be updated, rather than the entire language model. The financial services language model consists of candidate sub-models, each responsible for specific tasks and functions. Furthermore, a plug-in architecture allows for the insertion of new modules or functions, expanding the capabilities of the model without changing the existing architecture.

[0048] The financial services language model identifies keywords and intent in user-entered business questions and performs comprehensive semantic analysis based on the context to infer the fuzzy financial services items the user needs to process. A target sub-model associated with the fuzzy items is obtained from each candidate sub-model in the financial services language model. The target sub-model is invoked to interact with the user to obtain the target item to be processed. Exemplarily, the financial services language model includes at least one of an account management sub-model, a transfer and payment sub-model, a loan and credit sub-model, a credit card business sub-model, an investment and wealth management sub-model, and a risk management sub-model. The account management sub-model guides at least one of account opening, account inquiry, account change, and account cancellation. The transfer and payment sub-model guides at least one of intra-bank transfers, inter-bank transfers, cross-border payments, and mobile payments. The loan and credit sub-model guides at least one of loan applications, loan approvals, repayment management, and expectation management. The credit card business sub-model guides at least one of credit card applications, credit card activations, credit card spending, credit card repayments, and credit card points management. The investment and wealth management sub-model is used to guide at least one of the following: wealth management product recommendation business, fund trading business, stock trading business, and insurance product business. The risk management sub-model is used to guide at least one of the following: anti-fraud business, credit assessment business, market risk business, and operational risk business.

[0049] The large language model for banking business scenarios is modularly designed, the incremental training process is optimized, the banking business scenario modules are subdivided, and bank-specific data sets are used for training. This enables the financial service large language model to accurately understand users' financial needs and realize personalized business recommendations, greatly improving the efficiency and accuracy of users' confirmation of business transactions.

[0050] In an optional embodiment, the basic selection probability of each candidate machine is determined based on the target business item, the user's historical behavior data and the attribute information of each candidate machine, including: feature encoding the target business item, and the historical operation records, historical business processing records, and historical feedback evaluations in the historical behavior data to obtain a user feature vector; encoding the attribute information of each candidate machine to obtain candidate machine characteristics; wherein the attribute information includes at least one of the business scope, transaction limit range or user authority level; and performing similarity calculation on the user feature vector and the candidate machine characteristics to obtain the basic selection probability of each candidate machine.

[0051] The similarity between the user feature vector and the candidate tool features can be calculated using matrix decomposition methods, such as the alternating least squares method, as shown in the following formulas (1.1) and (1.2).

[0052] R≈PQ T (1.1)

[0053]

[0054] Where R is the user-candidate machine rating matrix, P is the user feature matrix with size m×k, where m is the number of users and k is the number of latent features, and Q is the candidate machine feature matrix with size n×k, where n is the number of candidate machines and k is the number of latent features. is the prediction result of user u for candidate machine j, that is, S j,pref , which is the basic selection probability of candidate machine j.

[0055] By constructing a user feature vector and the characteristics of each candidate machine, and calculating their similarity, we can determine the basic selection probability for each candidate machine. This accurately matches user needs with machine characteristics, providing users with more tailored candidate machine options, improving business processing efficiency and user satisfaction.

[0056] Figure 2 This is the second flow chart of a method for guiding financial services provided by an embodiment of the present invention. This embodiment is optimized and improved on the basis of the above embodiment. Figure 2 As shown, the method includes:

[0057] S210: Use the financial services big language model to analyze the business questions input by the user and obtain target business items to be processed.

[0058] S220 : Determine a basic selection probability of each candidate device based on the target service item, the user's historical behavior data, and the attribute information of each candidate device.

[0059] S230 : Determine the shortest path length and estimated arrival time between the candidate machine and the current position of the candidate machine, and determine a position correction probability of the candidate machine based on the shortest path length and estimated arrival time.

[0060] Specifically, after obtaining the user's consent, the user's current location is obtained through intelligent positioning technology, such as GPS positioning, IP address positioning, or WiFi positioning. The location of the candidate equipment is obtained from the bank's database or map service provider, and the location of the candidate equipment is integrated into the equipment information database. The distance between the user and each candidate equipment is calculated using Geographic Information System (GIS) technology, or the user and the candidate equipment are grouped using a geographic location clustering algorithm to determine the shortest path length between the user and the candidate equipment and the estimated time the user will arrive at the candidate equipment. The position correction probability of the candidate equipment is determined based on the shortest path length and the estimated arrival time, as shown in the following formula (2).

[0061]

[0062] Among them, d shortest represents the shortest path length from the user's current location to the candidate machine j, t expected is the estimated arrival time from the user’s current location to the candidate machine j, w1 and w2 are preset weights, S j,geo represents the position correction probability of candidate tool j.

[0063] S240: Determine the queuing correction probability of the candidate equipment according to the current business handling status of the candidate equipment.

[0064] The current business handling status of the candidate machine includes the queuing status of the candidate machine and the duration of each business that the candidate machine has to handle. The current business handling status of the candidate machine will affect the waiting time when the user arrives at the candidate machine. Based on the current business handling status of the candidate machine, the queuing correction probability of the candidate machine is determined.

[0065] In an optional embodiment, the queuing correction probability of the candidate equipment is determined based on the current business processing status of the candidate equipment, including: parsing the current business processing status of the candidate equipment to obtain the number of existing queuers of the candidate equipment, the estimated business processing time of the queuers, the total processing time of the currently processing business, and the processed time of the currently processing business; and determining the queuing correction probability of the candidate equipment based on the number of existing queuers, the estimated business processing time, the total processing time, and the processed time.

[0066] Monitoring devices or sensors installed on candidate machines collect real-time data on the machines' current business processing status. These devices, including at least one of cameras, infrared sensors, and pressure sensors, capture key information such as machine usage status, user behavior, and queue status. The collected data is transmitted via wired or wireless means to the digital twin financial service platform. Digital twin technology is used to create a virtual model corresponding to the physical machine, enabling simulation and analysis of the machine's operating status in a virtual environment. The data is cleaned, consolidated, and formatted to ensure accuracy and consistency.

[0067] The database is retrieved in real time to obtain the current business processing status of each candidate machine. Using machine learning, deep learning, and / or statistical analysis algorithms and techniques to explore patterns and trends in the data, the current business processing status of the candidate machine is analyzed to obtain information such as the number of existing queues for the candidate machine, the estimated business processing time for the queued customers, the total processing time of the current business, and the processing time of the current business, in order to assess the busyness of the candidate machine. By analyzing historical data and real-time monitoring, the queue status of each candidate machine is predicted. The queue correction probability S of candidate machine j is j,queue The calculation process is shown in the following formula (3).

[0068]

[0069] Where N is the number of people in the queue, t i,estimated is the expected business processing time of the i-th queuer, t current_total is the total processing time of the business currently being processed, t current_processed It is the processing time of the business currently being processed (counted from the start of processing), and ∈ is a small positive number used to prevent the denominator from being zero.

[0070] S250: Correct the basic selection probability using the position correction probability and the queue correction probability to obtain a target selection probability for the candidate machine.

[0071] S260 , sorting the candidate machines according to the target selection probability, so that the user can select a target machine from the candidate machines.

[0072] By determining the shortest path length between the user's current location and the candidate machines and the user's estimated arrival time at each candidate machine, the location correction probability of the candidate machines is determined. This allows for a more reasonable assessment of the accessibility of the candidate machines, improving the accuracy of machine recommendations and user convenience. The queue correction probability of the candidate machines is determined based on the current business status of the candidate machines. By correcting the basic selection probability of each candidate machine based on the queue modification positive probability, the actual usage efficiency and user experience of the candidate machines are more accurately reflected, thereby optimizing resource allocation, reducing user wait time, and improving overall service quality and user satisfaction.

[0073] S270: Guide the user to the target machine to handle the target business item.

[0074] This embodiment of the present invention uses both location-corrected and queue-corrected probabilities to modify the base selection probability. This approach comprehensively considers both the user's accessibility to the candidate machine and the machine's current busyness, resulting in a more accurate target selection probability. Candidate machines are ranked based on the target selection probability, providing users with machine selection recommendations that are more tailored to their actual needs, improving user satisfaction and selection efficiency.

[0075] In an optional embodiment, it also includes: extracting the user's business handling habits, time preferences and frequented locations from the historical behavior data; determining the behavior habit correction probability of the candidate device based on the current time period, the business handling habits, the time preferences and the frequented locations; and using the behavior habit correction probability to correct the target selection probability to obtain a corrected target selection probability.

[0076] Analyze users' business handling habits in different time periods, such as differences in behavior on weekdays and weekends, and their selection habits during morning and evening peaks. Identify the outlets or machine locations that users frequently visit, as well as their activity patterns within specific areas. Based on the user's historical behavior data, determine the behavioral habit correction probability of the candidate machine; use the behavioral habit correction probability to correct the target selection probability to obtain the corrected target selection probability. Specifically, use an autoregressive model (AR model) to determine the behavioral habit correction probability of the candidate machine. The calculation process of the behavioral habit correction probability is shown in the following formula (4).

[0077]

[0078] Among them, x t is the user's behavioral feature value at time t, and is also the probability of behavioral habit correction S j,habit , c is a constant term, φ i is the autoregressive coefficient, p is the lag order, v t is the noise error.

[0079] The target selection probability of each candidate machine is determined by combining multiple factors such as geographic location, queuing conditions, user preferences, and behavioral habits. The candidate machines are ranked and recommended to the user based on their target selection probabilities. The recommendation results are dynamically adjusted based on the user's real-time location and current needs to ensure that the recommended machines meet both user preferences and current business needs. During the recommendation process, multiple optimization objectives are considered simultaneously, such as minimizing user waiting time, maximizing user satisfaction, and balancing the utilization rate of each machine. The target selection probability of each candidate machine is calculated using a weighted average method, as shown in the following formula (5).

[0080] S j =w1S j,geo +w2S j,queue +w3S j,pref +w4S j,habit (5)

[0081] Among them, S j is the corrected target selection probability of candidate machine j, S j,geo is the position correction probability of candidate tool j, S j,queue is the queue correction probability of candidate machine j, S j,pref is the basic selection probability of candidate machine j, S j,habit is the modified probability of the behavior of candidate machine j, w1, w2, w3, and w4 are the weights of each probability, and w1 + w2 + w3 + w4 = 1. The candidate machines are sorted according to the modified target selection probability, allowing the user to select the target machine from among the candidate machines.

[0082] By correcting the target selection probability of each candidate machine through user historical behavior data, the ranking results of candidate machines can be optimized more personalized and accurately according to user preferences, helping users quickly lock in target machines and improve service experience and efficiency.

[0083] Figure 3 This is a schematic diagram of the structure of a guidance device for financial services provided by an embodiment of the present invention. Figure 3 As shown, the device includes:

[0084] The target business item module 310 is used to analyze the business question input by the user using the financial services language model to obtain the target business item to be processed;

[0085] A basic selection probability module 320 is configured to determine a basic selection probability for each candidate device based on the target service item, the user's historical behavior data, and the attribute information of each candidate device;

[0086] The target selection probability module 330 is configured to modify the basic selection probability based on the current business status of each candidate machine, the location of the candidate machine, and the current location of the user to obtain a target selection probability for each candidate machine, and to sort the candidate machines according to the target selection probability so that the user can select a target machine from the candidate machines.

[0087] The guidance module 340 is used to guide the user to go to the target machine to handle the target business item.

[0088] The financial service guidance device provided in the embodiment of the present invention can execute the financial service guidance method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0089] Optionally, the target business item module is specifically used to: use the financial service big language model to process the business problem input by the user to obtain the fuzzy business item to be processed, and obtain the target sub-model associated with the fuzzy business item from each candidate sub-model in the financial service big language model; call the target sub-model to interact with the user to obtain the target business item to be processed.

[0090] Optionally, the basic selection probability module is specifically used to: perform feature encoding on the target business item, as well as historical operation records, historical business processing records, and historical feedback evaluations in the historical behavior data to obtain a user feature vector; encode the attribute information of each candidate machine to obtain candidate machine features; wherein the attribute information includes at least one of the business scope, transaction limit range, or user authority level; perform similarity calculation on the user feature vector and the candidate machine features to obtain a basic selection probability for each candidate machine.

[0091] Optionally, the target selection probability module includes:

[0092] a position correction unit, configured to determine a shortest path length and an estimated arrival time between the candidate machine and the candidate machine based on the position of the candidate machine and the current position, and to determine a position correction probability of the candidate machine based on the shortest path length and the estimated arrival time;

[0093] A queuing correction unit, configured to determine a queuing correction probability of a candidate machine according to a current business handling status of the candidate machine;

[0094] The target selection probability unit is used to correct the basic selection probability by using the position correction probability and the queue correction probability to obtain the target selection probability of the candidate machine.

[0095] Optionally, the queue correction unit is specifically used to: analyze the current business processing status of the candidate machine to obtain the number of existing queuers, the estimated business processing time of the queuers, the total processing time of the currently processed business and the processed time of the currently processed business of the candidate machine; determine the queue correction probability of the candidate machine based on the number of existing queuers, the estimated business processing time, the total processing time and the processed time.

[0096] Optionally, it also includes a behavior habit correction module, which is used to extract the user's business handling habits, time preferences and frequented locations from the historical behavior data; determine the behavior habit correction probability of the candidate machine based on the current time period, the business handling habits, the time preferences and the frequented locations; and use the behavior habit correction probability to correct the target selection probability to obtain a corrected target selection probability.

[0097] The financial service business guidance device further described can also execute the financial service business guidance method provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0098] According to an embodiment of the present invention, the present invention further provides an electronic device, a readable storage medium and a computer program product.

[0099] Figure 4 A schematic diagram of the structure of an electronic device 40 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0100] like Figure 4As shown, the electronic device 40 includes at least one processor 41 and a memory, such as a read-only memory (ROM) 42, a random access memory (RAM) 43, etc., which is communicatively connected to the at least one processor 41. The memory stores a computer program that can be executed by the at least one processor, and the processor 41 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 42 or the computer program loaded from the storage unit 48 into the random access memory (RAM) 43. Various programs and data required for the operation of the electronic device 40 can also be stored in the RAM 43. The processor 41, ROM 42, and RAM 43 are connected to each other via a bus 44. An input / output (I / O) interface 45 is also connected to the bus 44.

[0101] Multiple components in the electronic device 40 are connected to the I / O interface 45, including an input unit 46, such as a keyboard, a mouse, etc.; an output unit 47, such as various types of displays, speakers, etc.; a storage unit 48, such as a magnetic disk, an optical disk, etc.; and a communication unit 49, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 49 allows the electronic device 40 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0102] Processor 41 can be any general-purpose and / or specialized processing component with processing and computing capabilities. Some examples of processor 41 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any other suitable processors, controllers, microcontrollers, etc. Processor 41 executes the various methods and processes described above, such as the bootstrapping method for financial services services.

[0103] In some embodiments, the bootstrapping method for a financial services service may be implemented as a computer program tangibly embodied in a computer-readable storage medium, such as storage unit 48. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 40 via ROM 42 and / or communication unit 49. When the computer program is loaded into RAM 43 and executed by processor 41, one or more steps of the bootstrapping method for a financial services service described above may be performed. Alternatively, in other embodiments, processor 41 may be configured to execute the bootstrapping method for a financial services service via any other suitable means (e.g., via firmware).

[0104] Various embodiments of the systems and techniques described herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chip systems (SOCs), programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0105] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when the computer program is executed by the processor, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The computer program may be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0106] In the context of the present invention, computer-readable storage media can be tangible media that can contain or store a computer program for use with an instruction execution system, device or equipment or used in combination with an instruction execution system, device or equipment. Computer-readable storage media can include but are not limited to electronic, magnetic, optical, electromagnetic, infrared or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, computer-readable storage media can be machine-readable signal media. More specific examples of machine-readable storage media can include electrical connections based on one or more lines, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0107] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0108] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, 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: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0109] A computing system may include clients and servers. The clients and servers are typically remote from each other and typically interact via a communication network. This client-server relationship arises through computer programs running on the respective computers, creating a client-server relationship. The server may be a cloud server, also known as a cloud computing server or cloud host. This server is a hosting product within the cloud computing service ecosystem that addresses the management difficulties and limited scalability of traditional physical hosting and VPS services.

[0110] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in the present invention can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of the present invention can be achieved. This is not limited herein.

[0111] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A method for guiding financial services, characterized in that: include: Use the financial services language model to analyze the business questions input by users and obtain the target business items to be handled; Determining a basic selection probability for each candidate device based on the target service item, the user's historical behavior data, and attribute information of each candidate device; The basic selection probability is modified based on the current business processing status of each candidate machine, the location of the candidate machine, and the current location of the user, including: determining the shortest path length and estimated arrival time between the user and the candidate machine based on the location of the candidate machine and the current location, and determining the location modification probability of the candidate machine based on the shortest path length and estimated arrival time; determining the queue modification probability of the candidate machine based on the current business processing status of the candidate machine; and modifying the basic selection probability using the location modification probability and the queue modification probability to obtain a target selection probability for each candidate machine; sorting the candidate machines according to the target selection probability, so that the user can select a target machine from the candidate machines; Guide the user to the target machine to handle the target business item.

2. The method according to claim 1, characterized in that The financial services large language model is used to analyze the business questions input by the user to obtain the target business items to be handled, including: Using the financial services large language model to process the business question input by the user to obtain a fuzzy business item to be processed, and obtaining a target sub-model associated with the fuzzy business item from each candidate sub-model in the financial services large language model; The target sub-model is called to interact with the user to obtain the target business item to be processed.

3. The method according to claim 1, characterized in that Determine the basic selection probability of each candidate device based on the target service item, the user's historical behavior data, and the attribute information of each candidate device, including: Performing feature coding on the target business item, and the historical operation records, historical business handling records, and historical feedback evaluations in the historical behavior data to obtain a user feature vector; Encoding attribute information of each candidate device to obtain characteristics of the candidate device; wherein the attribute information includes at least one of a business scope, a transaction limit range, or a user authority level; A similarity calculation is performed on the user feature vector and the candidate machine features to obtain a basic selection probability of each candidate machine.

4. The method according to claim 1, wherein The determining of the queuing correction probability of the candidate tool mode according to the current business handling status of the candidate tool includes: Analyze the current business processing status of the candidate equipment to obtain the number of existing queues for the candidate equipment, the estimated business processing time of the queues, the total processing time of the currently processed business, and the processing time of the currently processed business; The queue correction probability of the candidate tool is determined according to the number of existing queue members, the estimated business processing time, the total processing time and the processed time.

5. The method according to claim 1, wherein Also includes: Extracting the user's business handling habits, time preferences, and frequently visited locations from the historical behavior data; Determining a behavior habit modification probability of the candidate device based on the current time period, the business handling habits, the time preference, and the frequently visited locations; The target selection probability is corrected using the behavior habit correction probability to obtain a corrected target selection probability.

6. A guidance device for financial services, characterized in that: The device comprises: The target business item module is used to analyze the business questions input by users using the financial services large language model to obtain the target business items to be handled; A basic selection probability module is used to determine the basic selection probability of each candidate device based on the target service item, the user's historical behavior data and the attribute information of each candidate device; a target selection probability module, configured to modify the basic selection probability based on the current business handling status of each candidate machine, the location of the candidate machine, and the current location of the user, to obtain a target selection probability for each candidate machine, and to sort the candidate machines according to the target selection probability so that the user can select a target machine from among the candidate machines; A guidance module, used to guide the user to the target machine to handle the target business item; Wherein, the target selection probability module includes: a position correction unit, configured to determine a shortest path length and an estimated arrival time between the user and the candidate machine based on the position of the candidate machine and the current position, and to determine a position correction probability of the candidate machine based on the shortest path length and the estimated arrival time; A queuing correction unit, configured to determine a queuing correction probability of a candidate machine according to a current business handling status of the candidate machine; The target selection probability unit is used to correct the basic selection probability by using the position correction probability and the queue correction probability to obtain the target selection probability of each candidate machine.

7. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively coupled to the at least one processor; The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the method for guiding financial services according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for guiding a financial service business according to any one of claims 1 to 5 when executed.

9. A computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements the method for guiding a financial services service according to any one of claims 1 to 5.

Images