Business object recommendation method, device, computer device and readable storage medium

Abstract

The present application relates to a method, apparatus, computer device and readable storage medium for recommending business objects. The method comprises: obtaining business characteristic data corresponding to a resource allocation account set, and establishing a business selection map domain space corresponding to the resource allocation account set based on the business characteristic data; the business selection map domain space comprises at least two business object points; based on each business object point, at least two candidate business object point selection paths are constructed; each candidate business object point selection path has a corresponding selection probability; a target business object point selection path is determined in each candidate business object point selection path; and the target business object represented by the target business object point selection path is recommended to the target resource allocation account in the resource allocation account set. The use of this method can save time in searching for business recommendations, and can also provide an effective basis for recommending resource allocation account products, thereby improving the accuracy of business recommendations.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method, apparatus, computer device, and readable storage medium for recommending a business object. Background Art

[0002] With the development of computer technology, intelligent business recommendation technology has emerged. This technology is a software tool and technical method that can recommend useful products to users. These recommendations are used in various decision-making processes, such as purchasing items, selecting music, and browsing news. Intelligent business recommendation technology is also an advanced business intelligence platform based on massive data mining, providing customers with personalized information services and decision support.

[0003] Traditionally, intelligent service recommendation technology collects information from multiple customers, categorizes it based on big data, and then recommends relevant services to those customers based on their needs. However, since big data categorization only targets customers of the same type, detailed analysis of each individual customer is not possible. Consequently, service recommendations for specific customers often deviate from the norm, failing to accurately recommend suitable services. Consequently, service recommendations often fail to meet pre-defined expectations. Summary of the Invention

[0004] Based on this, it is necessary to provide a business object recommendation method, apparatus, computer device, computer-readable storage medium and computer program product that can improve the accuracy of business recommendations for target resource allocation accounts in response to the above technical problems.

[0005] In a first aspect, the present application provides a business object recommendation method. The method comprises: obtaining business characteristic data corresponding to a resource allocation account set, and establishing a business selection map domain space corresponding to the resource allocation account set based on the business characteristic data; the business selection map domain space includes at least two business object points; the business object points are used to represent candidate business objects to be selected; at least two candidate business object point selection paths are constructed based on each of the business object points; each candidate business object point selection path has a corresponding selection probability; the selection probability is determined based on the directional probability between two adjacent business object points in the corresponding candidate business object point selection path; the directional probability is used to indicate a path formed by an intelligent agent passing through two adjacent business object points; determining a target business object point selection path in each of the candidate business object point selection paths; the selection probability corresponding to the target business object point selection path is greater than the selection probability corresponding to other business object point selection paths; the other business object point selection paths are candidate business object point selection paths in each of the candidate business object point selection paths other than the target business object point selection path; and recommending the target business object represented by the target business object point selection path to the target resource allocation account in the resource allocation account set.

[0006] In the second aspect, the present application also provides a business object recommendation device. The device includes: a map domain space construction module, which is used to obtain the business characteristic data corresponding to the resource allocation account set, and establish a business selection map domain space corresponding to the resource allocation account set based on the business characteristic data; the business selection map domain space includes at least two business object points; the business object points are used to represent the candidate business objects to be selected; a candidate path construction module, which is used to construct at least two candidate business object point selection paths based on each of the business object points; each of the candidate business object point selection paths has a corresponding selection probability; the selection probability is determined based on the direction probability between the two adjacent business object points in the corresponding candidate business object point selection path. ; The direction probability is used to indicate the path formed by the intelligent agent passing through two adjacent business object points; the target path determination module is used to determine the target business object point selection path in each of the candidate business object point selection paths; the selection probability corresponding to the target business object point selection path is greater than the selection probabilities corresponding to other business object point selection paths; the other business object point selection paths are the candidate business object point selection paths in each of the candidate business object point selection paths except the target business object point selection path; the business recommendation module is used to recommend the target business object represented by the target business object point selection path to the target resource allocation account in the resource allocation account set.

[0007] In a third aspect, the present application also provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program: obtaining business characteristic data corresponding to the resource allocation account set, and establishing a business selection map domain space corresponding to the resource allocation account set based on the business characteristic data; the business selection map domain space includes at least two business object points; the business object points are used to represent candidate business objects to be selected; at least two candidate business object point selection paths are constructed based on each of the business object points; each candidate business object point selection path has a corresponding selection probability; the selection probability is determined based on the corresponding candidate business object point selection path. The direction probability between two adjacent business object points in the path is determined; the direction probability is used to indicate the path formed by the intelligent agent passing through two adjacent business object points; the target business object point selection path is determined in each of the candidate business object point selection paths; the selection probability corresponding to the target business object point selection path is greater than the selection probability corresponding to other business object point selection paths; the other business object point selection paths are the candidate business object point selection paths in each of the candidate business object point selection paths except the target business object point selection path; the target business object represented by the target business object point selection path is recommended to the target resource allocation account in the resource allocation account set.

[0008] In a fourth aspect, the present application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, and when the computer program is executed by a processor, the following steps are implemented: obtaining business characteristic data corresponding to the resource allocation account set, and establishing a business selection map domain space corresponding to the resource allocation account set based on the business characteristic data; the business selection map domain space includes at least two business object points; the business object points are used to represent candidate business objects to be selected; at least two candidate business object point selection paths are constructed based on each of the business object points; each of the candidate business object point selection paths has a corresponding selection probability; the selection probability is determined based on the probability of the two adjacent business object points in the corresponding candidate business object point selection path. The direction probability between the business object points is determined; the direction probability is used to indicate the path formed by the intelligent agent passing through two adjacent business object points; the target business object point selection path is determined in each of the candidate business object point selection paths; the selection probability corresponding to the target business object point selection path is greater than the selection probability corresponding to other business object point selection paths; the other business object point selection paths are the candidate business object point selection paths in each of the candidate business object point selection paths except the target business object point selection path; the target business object represented by the target business object point selection path is recommended to the target resource allocation account in the resource allocation account set.

[0009] In a fifth aspect, the present application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, implements the following steps: obtaining business characteristic data corresponding to a resource allocation account set, and establishing a business selection map domain space corresponding to the resource allocation account set based on the business characteristic data; the business selection map domain space includes at least two business object points; the business object points are used to represent candidate business objects to be selected; at least two candidate business object point selection paths are constructed based on each of the business object points; each candidate business object point selection path has a corresponding selection probability; the selection probability is determined based on a directional probability between two adjacent business object points in the corresponding candidate business object point selection path; the directional probability is used to indicate a path formed by an intelligent agent passing through two adjacent business object points; determining a target business object point selection path in each of the candidate business object point selection paths; the selection probability corresponding to the target business object point selection path is greater than the selection probabilities corresponding to other business object point selection paths; the other business object point selection paths are candidate business object point selection paths in each of the candidate business object point selection paths other than the target business object point selection path; and recommending the target business object represented by the target business object point selection path to the target resource allocation account in the resource allocation account set.

[0010] The above-mentioned business object recommendation method, device, computer equipment, storage medium and computer program product obtain business characteristic data corresponding to the resource allocation account set and establish a business selection map domain space corresponding to the resource allocation account set based on the business characteristic data; the business selection map domain space includes at least two business object points; the business object point is used to represent the candidate business object to be selected; at least two candidate business object point selection paths are constructed based on each business object point; each candidate business object point selection path has a corresponding selection probability; the selection probability is determined based on the direction probability between two adjacent business object points in the corresponding candidate business object point selection path; the direction probability is used to indicate the path composed of two adjacent business object points passed by the intelligent agent; the target business object point selection path is determined in each candidate business object point selection path; the selection probability corresponding to the target business object point selection path is greater than the selection probability corresponding to other business object point selection paths; the other business object point selection paths are the candidate business object point selection paths in each candidate business object point selection path except the target business object point selection path; the target business object represented by the target business object point selection path is recommended to the target resource allocation account in the resource allocation account set.

[0011] By abstracting customer information and resource allocation account information, constructing an abstract environment map model, and using an optimization algorithm to search in the map, the heuristic value algorithm and pheromone update method in the optimization algorithm are improved, and the transfer probability is calculated to select a better credit card product. Through continuous iterative search, the resource allocation account product with the highest selection probability under the existing conditions is obtained. It can effectively consider the relevant information of the customer and the various attribute information of the resource allocation account, and use the improved optimization algorithm to search for the resource allocation account product with the highest selection probability. It not only saves time, but also provides an effective basis for the recommendation of resource allocation account products, thereby improving the accuracy of business recommendations. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 An application environment diagram of a business object recommendation method in one embodiment;

[0013] Figure 2 A schematic diagram of a flow chart of a business object recommendation method in one embodiment;

[0014] Figure 3 A schematic diagram of a flow chart of a method for determining a target business object and selecting a path in one embodiment;

[0015] Figure 4 Schematic diagram of a flow chart of a method for updating global pheromones in one embodiment;

[0016] Figure 5 1 is a flow chart of a method for updating local pheromones in one embodiment;

[0017] Figure 6 A schematic diagram of a process for constructing a heuristic function method in one embodiment;

[0018] Figure 7 A schematic flow chart of a method for constructing a heuristic function in another embodiment;

[0019] Figure 8 A schematic diagram of a process for setting parameters of a business abstract environment model in one embodiment;

[0020] Figure 9 A flowchart of a method for selecting a resource allocation account product with the highest probability obtained in one embodiment;

[0021] Figure 10 A flowchart of a business object recommendation method according to another embodiment;

[0022] Figure 11 It is a structural block diagram of a business object recommendation device in one embodiment;

[0023] Figure 12FIG. 1 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0025] The business object recommendation method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown, terminal 102 acquires data, and server 104 receives the data from terminal 102 in response to instructions from terminal 102 and performs calculations on the acquired data. Server 104 transmits the calculation results back to terminal 102, which then displays them. Terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated with server 104 or placed in the cloud or on other network servers. The server 104 obtains the business characteristic data corresponding to the resource allocation account set from the terminal 102, and establishes a business selection map domain space corresponding to the resource allocation account set based on the business characteristic data; the business selection map domain space includes at least two business object points; the business object point is used to represent the candidate business object to be selected; based on each business object point, at least two candidate business object point selection paths are constructed; each candidate business object point selection path has a corresponding selection probability; the selection probability is determined based on the direction probability between two adjacent business object points in the corresponding candidate business object point selection path; the direction probability is used to indicate the path composed of two adjacent business object points passed by the intelligent agent; the target business object point selection path is determined in each candidate business object point selection path; the selection probability corresponding to the target business object point selection path is greater than the selection probability corresponding to other business object point selection paths; the other business object point selection paths are the candidate business object point selection paths in each candidate business object point selection path except the target business object point selection path; the target business object represented by the target business object point selection path is recommended to the target resource allocation account in the resource allocation account set. Terminal 102 may include, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices may include smart speakers, smart TVs, smart air conditioners, and smart car devices. Portable wearable devices may include smart watches, smart bracelets, and head-mounted devices. Server 104 may be implemented as a standalone server or a server cluster consisting of multiple servers.

[0026] In one embodiment, Figure 2 As shown, a business object recommendation method is provided, which is applied to Figure 1 The following steps are used as an example to illustrate the server in the example:

[0027] Step 202: Acquire business characteristic data corresponding to the resource allocation account set, and establish a business selection map domain space corresponding to the resource allocation account set based on the business characteristic data.

[0028] Among them, the resource allocation account set can be a group consisting of at least two resource allocation accounts, wherein the resource allocation account can be a personal account representing the interaction of social or natural resources on a public resource platform, and the account records the information of the interaction and the information of the account owner.

[0029] The business characteristic data may be data related to resource interaction, such as personal information required for resource interaction, interaction information generated by resource interaction, computer messages that need to be tracked after resource interaction, etc.

[0030] The business selection map domain space may be an abstract map representing the specific positions between various business object points and the relative positions between various business object points. The business selection map domain space is composed of a business abstract environment model.

[0031] Among them, the business selection point can be an abstract point representing different businesses in the resource allocation account, and a location point on the map used by the intelligent agent to perform business optimization.

[0032] Specifically, the server responds to the terminal's instruction, obtains the business characteristic data corresponding to the resource allocation account set from the terminal, and stores the obtained business characteristic data corresponding to the resource allocation account set in a storage unit. When the server needs to process a data record corresponding to any information in the business characteristic data corresponding to the resource allocation account set, it retrieves the data from the storage unit to a volatile storage resource for calculation by the central processor. The business characteristic data can be input as a single data item or as multiple data items simultaneously.

[0033] Based on the agent-related attributes in the customer information abstraction algorithm, and based on the map environment attributes in the resource allocation account information abstraction algorithm. The attributes of the resource allocation account include the resource allocation account type, installment payment, preferential activities, consumption limit, etc. The resource allocation account attribute information is shown in the following table. Figure 9As shown, different points on the map represent resource allocation accounts with different attributes, randomly distributed across the map. This serves as the business selection map domain space for product selection for resource allocation accounts. Customer information, including whether they are new customers, age, occupation, education level, income, expenditure, and other basic information, as well as available resource allocation account types and spending limits, key consumer expenditure characteristics, and social attribute information, serves as parameters and constraints for agent movement. The business selection map domain space includes at least two business object points; these points represent candidate business objects to be selected.

[0034]

[0035] Step 204: Construct at least two candidate business object point selection paths based on each business object point.

[0036] Among them, the candidate business object point selection path can be a selection path constructed in the business selection map domain space based on at least two business object points. When the starting point and end point of different candidate business object point selection paths are the same, the intermediate business selection points constituting the path can be different.

[0037] Specifically, the business object point corresponding to the starting point and the business object point corresponding to the end point of the business selection map domain space agent are determined, and based on each business object point in the business selection map domain space, a candidate business object point selection path in the business selection map domain space is constructed. Because there are multiple different business object points between the business object point corresponding to the starting point and the business object point corresponding to the end point, the constructed candidate business object point selection path is closely related to the business object point. Each candidate business object point selection path has a corresponding selection probability; the selection probability is determined based on the directional probability between two adjacent business object points in the corresponding candidate business object point selection path; the directional probability is used to indicate the path formed by the agent passing through two adjacent business object points.

[0038] Among them, for the agent's location selection, the agent selects the next node through the roulette method and repeats until it reaches the target point. The probability formula for selecting the next point is as follows:

[0039]

[0040]

[0041]

[0042] Among them, τ ij is the pheromone trajectory from grid i to grid j, and η ijis the heuristic value information from grid i to grid j. α is the weighting factor of pheromone concentration, and β is the weighting factor of heuristic value information, which determine the influence of pheromone concentration and heuristic value on node selection. ij is the difference in resource allocation account attributes between grid i and grid j. (x i ,y i ) and (x j ,y j ) is the attribute value of grid i and grid j. allow k is the set of grids that the agent can choose when it is at grid i.

[0043] After each iteration of the agent's search from the starting point to the target point, the pheromone is updated based on the length of the planned path. Initially, the agent has no pheromone value to refer to, so the path selection is random. For shorter paths, where the agent has only a short time to pass through, the pheromone will not fully evaporate, causing the next agent to explore the nodes on the path. This increases the pheromone concentration along the path, allowing subsequent agents to more likely ignore unqualified tracks and improve their coverage efficiency to find shorter paths. After n moments, the agent completes a cycle, and the pheromone update formula is as follows:

[0044]

[0045] Where m is the number of agents in each agent. ρ is the pheromone evaporation rate. Δτ ij represents the value of pheromone left by the kth agent in the path from grid i to grid j, Δτ ij The calculation method is as follows:

[0046]

[0047] Where Q1 is a constant. k (t) is the length of the path found by the kth agent.

[0048] Step 206: Determine a target business object point selection path among the candidate business object point selection paths.

[0049] The target business object point selection path may be a path selected by an intelligent agent, and then a path with the greatest business relevance is determined from among candidate business object point selection paths.

[0050] Specifically, the agent searches for candidate business object point selection paths in the business selection map domain space, and determines the target business object point selection path corresponding to the business recommendation for the resource allocation account from the candidate business object point selection paths. The target business object point selection path has a higher selection probability than the other business object point selection paths; the other business object point selection paths are the candidate business object point selection paths other than the target business object point selection path.

[0051] Among them, the search strategy for each agent. The heuristic function is essential in guiding the search process of each agent and plays an important role in the algorithm to quickly, effectively and accurately plan feasible paths. When the agent moves from the current grid to the next grid, the selection probability of each candidate grid within the visible area can be calculated according to the heuristic function. The heuristic function can be represented by F(i, j, k), and its calculation process is as follows:

[0052]

[0053]

[0054]

[0055]

[0056] Where o is the current point, t is the candidate point, and d is the target point. M(i, j, k) is the resource allocation account benefit between the current agent's target point and the point it is currently located at, prompting the agent to choose a point with higher benefits. S(i, j, k) is a safety factor that encourages the agent to choose a feasible point. m is the number of selectable points in the agent's visual domain at the current point, and m′ is the number of unselectable points in the visual domain. N(i, j, k) is the resource allocation account benefit between the target point and the candidate point, prompting the agent to choose a point closer to the target point. w1, w2, and w3 are parameters representing the importance of each factor.

[0057] Among them, the resource allocation account probability calculation. Based on the heuristic value, pheromone value and resource allocation account income, the selection probability p(i′, j′, k′) of each candidate point (i′, j′, k′) in the visible area is calculated. This ranking allows the algorithm to prioritize nodes with high probabilities, thereby achieving the goal of high-speed search. The calculation formula for the selection probability p(i′, j′, k′) is as follows:

[0058]

[0059] Among them, τ is the pheromone value of grid (i′, j′, k′), the initial pheromone value set in the first search, F is the heuristic value of grid (i′, j′, k′), and N is the difference in resource allocation account benefits calculated for grid (i′, j′, k′) compared with the original position.

[0060] Among them, local pheromone update. During each agent's search, a local search algorithm is used to extract and update pheromones from the visited grid. This behavior is called local pheromone update. Its main purpose is to increase pheromones when the agent passes through a grid to transmit grid information to subsequent agents. The local pheromone update formula is as follows: τ is the pheromone value on the grid; ξ is the pheromone attenuation coefficient. For example, when agents in nature find an optimal foraging path, once other agents follow it to find food, they will continue to strengthen the scent. Conversely, if no agent passes through the path again, the pheromone will gradually dissipate and decay. The attenuation coefficient can be a manually set parameter, generally a number between 0 and 1, and is set to 0.7 in this article.

[0061] τ=(1-ζ)*τ

[0062] Among them, global pheromone update. Agents need to use global pheromone information, rather than just local pheromone information, to guide them in choosing the path that best meets their goal. Global update means that after an agent completes its entire search, it selects the path that best meets its goal and adds pheromones to each grid. This increases the pheromone concentration in each grid on the optimal path and prepares the search of each agent in the next iteration. The global pheromone update formula is as follows: where τ is the pheromone value on the grid; ρ is the pheromone update coefficient; K is a constant; and length(n) is the length of the path traversed by the nth agent.

[0063]

[0064] The selectability of each grid in the map can be dynamically changed through local pheromone updates and global pheromone updates. When each agent selects a grid in each iteration, it can receive pheromones to understand whether the grid meets the path grid of the optimization target, thereby obtaining a planning path that better meets the target, such as Figure 9 As shown in FIG, the resource allocation account product with the highest selection probability is obtained through the resource allocation account product selection model based on each intelligent agent algorithm.

[0065] Step 208 : Recommend the target business object represented by the target business object point selection path to the target resource allocation account in the resource allocation account set.

[0066] Specifically, the target business object corresponding to the target business object point selection path obtained by the agent selection is recommended to the target resource allocation account in the resource allocation account set, and is recommended to the customer as a business object in the target resource allocation account. The flowchart of the overall business object recommendation method is as follows: Figure 10 shown.

[0067] In the above-mentioned business object recommendation method, business characteristic data corresponding to the resource allocation account set is obtained, and a business selection map domain space corresponding to the resource allocation account set is established based on the business characteristic data; the business selection map domain space includes at least two business object points; the business object point is used to represent the candidate business object to be selected; at least two candidate business object point selection paths are constructed based on each business object point; each candidate business object point selection path has a corresponding selection probability; the selection probability is determined based on the direction probability between two adjacent business object points in the corresponding candidate business object point selection path; the direction probability is used to indicate the path composed of two adjacent business object points passed by the intelligent agent; the target business object point selection path is determined in each candidate business object point selection path; the selection probability corresponding to the target business object point selection path is greater than the selection probability corresponding to other business object point selection paths; the other business object point selection paths are the candidate business object point selection paths in each candidate business object point selection path except the target business object point selection path; the target business object represented by the target business object point selection path is recommended to the target resource allocation account in the resource allocation account set.

[0068] By abstracting customer information and resource allocation account information, constructing an abstract environment map model, and using an optimization algorithm to search in the map, the heuristic value algorithm and pheromone update method in the optimization algorithm are improved, and the transfer probability is calculated to select a better credit card product. Through continuous iterative search, the resource allocation account product with the highest selection probability under the existing conditions is obtained. It can effectively consider the relevant information of the customer and the various attribute information of the resource allocation account, and use the improved optimization algorithm to search for the resource allocation account product with the highest selection probability. It not only saves time, but also provides an effective basis for the recommendation of resource allocation account products, thereby improving the accuracy of business recommendations.

[0069] In one embodiment, Figure 3 As shown, determining a target business object point selection path among each candidate business object point selection path includes:

[0070] Step 302: When each business object point has passability in any direction, the intelligent agent is controlled to perform a global search operation in the business selection map domain space and update the global pheromone corresponding to the business selection map domain space.

[0071] Among them, the global pheromone can be the information left in the business selection map domain space for guiding the path after the intelligent agent walks through the business selection map domain space and searches for the best path.

[0072] Specifically, when each business object point has passability in any direction, the search strategy for each agent is determined. The heuristic function is essential in guiding the search process of each agent and plays an important role in the algorithm's rapid, effective, and accurate planning of feasible paths. When the agent moves from the current grid to the next grid, the selection probability of each candidate grid within the visible area can be calculated based on the heuristic function. The heuristic function can be represented by F(i, j, k), and its calculation process is as follows:

[0073]

[0074]

[0075]

[0076]

[0077] Where o is the current point, t is the candidate point, and d is the target point. M(i, j, k) is the resource allocation account benefit between the current agent's target point and the point it is currently located at, prompting the agent to choose a point with higher benefits. S(i, j, k) is a safety factor that encourages the agent to choose a feasible point. m is the number of selectable points in the agent's visual domain at the current point, and m′ is the number of unselectable points in the visual domain. N(i, j, k) is the resource allocation account benefit between the target point and the candidate point, prompting the agent to choose a point closer to the target point. w1, w2, and w3 are parameters representing the importance of each factor.

[0078] Among them, the resource allocation account probability calculation. Based on the heuristic value, pheromone value and resource allocation account income, the selection probability p(i′, j′, k′) of each candidate point (i′, j′, k′) in the visible area is calculated. This ranking allows the algorithm to prioritize nodes with high probabilities, thereby achieving the goal of high-speed search. The calculation formula for the selection probability p(i′, j′, k′) is as follows:

[0079]

[0080] Among them, τ is the pheromone value of grid (i′, j′, k′), the initial pheromone value set in the first search, F is the heuristic value of grid (i′, j′, k′), and N is the difference in resource allocation account benefits calculated for grid (i′, j′, k′) compared with the original position.

[0081] Among them, local pheromone update. During each agent's search, a local search algorithm is used to extract and update pheromones from the visited grid. This behavior is called local pheromone update. Its main purpose is to increase pheromones when the agent passes through a grid to transmit grid information to subsequent agents. The local pheromone update formula is as follows: τ is the pheromone value on the grid; ξ is the pheromone attenuation coefficient. For example, when agents in nature find an optimal foraging path, once other agents follow it to find food, they will continue to strengthen the scent. Conversely, if no agent passes through the path again, the pheromone will gradually dissipate and decay. The attenuation coefficient can be a manually set parameter, generally a number between 0 and 1, and is set to 0.7 in this article.

[0082] τ=(1-ζ)*τ

[0083] Among them, global pheromone update. Agents need to use global pheromone information, rather than just local pheromone information, to guide them in choosing the path that best meets their goal. Global update means that after an agent completes its entire search, it selects the path that best meets its goal and adds pheromones to each grid. This increases the pheromone concentration in each grid on the optimal path and prepares the search of each agent in the next iteration. The global pheromone update formula is as follows: where τ is the pheromone value on the grid; ρ is the pheromone update coefficient; K is a constant; and length(n) is the length of the path traversed by the nth agent.

[0084]

[0085] Step 304, when the number of times the global search operation is performed does not reach the preset number of iterations, return to the step of obtaining the business characteristic data corresponding to the resource allocation account set, and establishing the business selection map domain space corresponding to the resource allocation account set based on the business characteristic data, until the number of global searches meets the preset number of iterations.

[0086] Specifically, when the number of times the global search operation is executed does not reach the preset number of iterations, the server returns to execute the instruction of "the server responds to the terminal, obtains the business characteristic data corresponding to the resource allocation account set from the terminal, and stores the obtained business characteristic data corresponding to the resource allocation account set in the storage unit. When the server needs to process the data record corresponding to any information in the business characteristic data corresponding to the resource allocation account set, it is retrieved from the storage unit to the volatile storage resource for the central processor to perform calculation. The business characteristic data can be input as a single data or as multiple data simultaneously.

[0087] Based on the agent-related attributes in the customer information abstraction algorithm, and based on the map environment attributes in the resource allocation account information abstraction algorithm. The attributes of the resource allocation account include the resource allocation account type, installment payment, preferential activities, consumption limit, etc. The resource allocation account attribute information is shown in the following table. Figure 9 As shown, different points on the map represent resource allocation accounts with different attributes, randomly distributed across the map, serving as the business selection map domain space for product selection for resource allocation accounts. Customer information, including whether they are new customers, age, occupation, education, income, expenditure, and other basic information, available resource allocation account types and spending limits, key consumer expenditure characteristics, and social attribute information, serves as parameters and constraints for agent movement. The business selection map domain space includes at least two business object points; these business object points are used to represent candidate business objects to be selected, and the process continues until the global search reaches a preset number of iterations.

[0088] Step 306 : When the number of times the global search operation is performed meets the preset number of iterations, a target business object point selection path is determined in the business selection map domain space according to the updated global pheromone.

[0089] Specifically, local pheromone updates and global pheromone updates can dynamically change the selectability of each grid in the map. When each agent selects a grid in each iteration, it can receive pheromones to understand whether the grid meets the path grid of the optimization target, thereby obtaining a planning path that better meets the target, such as Figure 9 As shown in FIG, the resource allocation account product with the highest selection probability is obtained through the resource allocation account product selection model based on each intelligent agent algorithm.

[0090] In this embodiment, by controlling the intelligent agent multiple times to perform global search operations in the business selection map domain space and updating the global pheromone, the target business object point selection path can be determined based on the prompts of the global pheromone, thereby improving the accuracy of business recommendations for the target resource allocation account.

[0091] In one embodiment, Figure 4 As shown in the figure, when each business object point has passability in any direction, the control agent performs a global search operation in the business selection map domain space and updates the global pheromone corresponding to the business selection map domain space, including:

[0092] Step 402 : Based on the fact that each business object point has passability in any direction, obtain the object point heuristic value corresponding to each business object point.

[0093] Specifically, when there is passability in any direction among the business object points, the heuristic function can be used to calculate the heuristic value. Search strategy for each agent. The heuristic function is indispensable in guiding the search process of each agent and plays an important role in the algorithm to plan feasible paths quickly, effectively and accurately. When the agent moves from the current grid to the next grid, the selection probability of each candidate grid within the visible area can be calculated according to the heuristic function. The heuristic function can be represented by F(i, j, k), and its calculation process is as follows:

[0094]

[0095]

[0096]

[0097]

[0098] Where o is the current point, t is the candidate point, and d is the target point. M(i, j, k) is the resource allocation account benefit between the current agent's target point and the point it is currently located at, prompting the agent to choose a point with higher benefits. S(i, j, k) is a safety factor that encourages the agent to choose a feasible point. m is the number of selectable points in the agent's visual domain at the current point, and m′ is the number of unselectable points in the visual domain. N(i, j, k) is the resource allocation account benefit between the target point and the candidate point, prompting the agent to choose a point closer to the target point. w1, w2, and w3 are parameters representing the importance of each factor.

[0099] Step 404: Based on the object point heuristic value, control the agent to move from the current business object point to the next business object point, and update the local pheromone corresponding to the current business object point.

[0100] The current business object point may be a business object point expressing where a certain intelligent agent is currently located in the business selection map domain space.

[0101] The next business object point may be a business object point to which an intelligent agent is ready to move in the business selection map domain space.

[0102] Among them, the local pheromone can be the information left by the intelligent agent when performing any step of search in the business point in the business selection map domain space.

[0103] Specifically, based on the heuristic value, pheromone value, and resource allocation account benefits, the selection probability p(i′, j′, k′) of each candidate point (i′, j′, k′) in the visible area is calculated. This ranking allows the algorithm to prioritize nodes with high probabilities, thereby achieving the goal of high-speed search. The selection probability of each direction in the visible area is calculated, and the importance of the search direction is estimated and sorted from large to small; the next moving position is selected based on the selection probability of each direction in the visible area and the local pheromone value is updated. The calculation formula for the direction selection probability p(i′, j′, k′) is as follows:

[0104]

[0105] Among them, τ is the pheromone value of grid (i′, j′, k′), the initial pheromone value set in the first search, F is the heuristic value of grid (i′, j′, k′), and N is the difference in resource allocation account benefits calculated for grid (i′, j′, k′) compared with the original position.

[0106] Among them, local pheromone update. During each agent's search, a local search algorithm is used to extract and update pheromones from the visited grid. This behavior is called local pheromone update. Its main purpose is to increase pheromones when the agent passes through a grid to transmit grid information to subsequent agents. The local pheromone update formula is as follows: τ is the pheromone value on the grid; ξ is the pheromone attenuation coefficient. For example, when agents in nature find an optimal foraging path, once other agents follow it to find food, they will continue to strengthen the scent. Conversely, if no agent passes through the path again, the pheromone will gradually dissipate and decay. The attenuation coefficient can be a manually set parameter, generally a number between 0 and 1, and is set to 0.7 in this article.

[0107] τ=(1-ζ)*τ

[0108] Step 406, when any agent fails to complete the search operation for the candidate business object point selection path, returns to the execution state where there is passability in any direction in each business object point, controls the agent to perform a global search operation in the business selection map domain space, and updates the global pheromone corresponding to the business selection map domain space until each agent completes the search operation for the candidate business object point selection path.

[0109] Specifically, if any agent has not completed the search operation for the candidate business object point selection path, it returns to execute "Step 302". For detailed process, please refer to the process description corresponding to Step 302 until each agent completes the search operation for the candidate business object point selection path.

[0110] Step 408: Based on the search operation of each agent on the candidate business object point selection path, the global pheromone corresponding to the business selection map domain space is updated.

[0111] Specifically, agents need to use global pheromone information, rather than just local pheromone information, to guide them in selecting the path that best meets their goal. Global updating involves selecting the path that best meets the goal and adding pheromones to each grid after an agent completes its search. This increases the pheromone concentration in each grid along the optimal path, preparing for the next agent's search in the next iteration. The global pheromone update formula is as follows: τ is the pheromone value on the grid; ρ is the pheromone update coefficient; K is a constant; and length(n) is the length of the path traversed by the nth agent.

[0112]

[0113] In this embodiment, by controlling all intelligent agents to perform full-area search operations in the business selection map domain space and updating the global pheromone, all intelligent agents can walk from the starting business object point to the ending business object point and leave corresponding local pheromones, effectively improving the strength of the global pheromone and providing evidence for the selection of the path for the target business object point.

[0114] In one embodiment, Figure 5 As shown, controlling the agent to move from the current business object point to the next business object point and updating the local pheromone corresponding to the current business object point includes:

[0115] Step 502: Obtain the direction selection probabilities corresponding to the directions of the current business object point, and sort the direction selection probabilities.

[0116] The direction selection probability may be the probability of the agent moving from the current business object point to each next business object point in the business selection map domain space.

[0117] Specifically, based on the heuristic value, pheromone value, and resource allocation account benefits, the selection probability p(i′, j′, k′) of each candidate point (i′, j′, k′) in the visible area is calculated. This ranking allows the algorithm to prioritize nodes with high probabilities, thereby achieving the goal of high-speed search. The selection probability of each direction in the visible area is calculated, and the importance of the search direction is estimated and sorted from large to small. The calculation formula for the direction selection probability p(i′, j′, k′) is as follows:

[0118]

[0119] Among them, τ is the pheromone value of grid (i′, j′, k′), the initial pheromone value set in the first search, F is the heuristic value of grid (i′, j′, k′), and N is the difference in resource allocation account benefits calculated for grid (i′, j′, k′) compared with the original position.

[0120] Step 504: Based on the sorting results corresponding to the direction selection probabilities, the agent is controlled to select a direction selection probability that satisfies a preset probability condition, moves from the current business object point to the next business object point, and updates the local pheromone corresponding to the business object point.

[0121] Specifically, based on the importance of the selection probability of each direction and sorting the results from large to small, the control agent selects the direction corresponding to the maximum direction selection probability, selects the next moving position and updates the local pheromone value.

[0122] Among them, local pheromone update. During each agent's search, a local search algorithm is used to extract and update pheromones from the visited grid. This behavior is called local pheromone update. Its main purpose is to increase pheromones when the agent passes through a grid to transmit grid information to subsequent agents. The local pheromone update formula is as follows: τ is the pheromone value on the grid; ξ is the pheromone attenuation coefficient. For example, when agents in nature find an optimal foraging path, once other agents follow it to find food, they will continue to strengthen the scent. Conversely, if no agent passes through the path again, the pheromone will gradually dissipate and decay. The attenuation coefficient can be a manually set parameter, generally a number between 0 and 1, and is set to 0.7 in this article.

[0123] τ=(1-ζ)*τ

[0124] In this embodiment, by sorting the direction selection probabilities corresponding to each direction, selecting the direction whose direction selection probability meets the preset conditions to control the movement of the intelligent agent, and updating the local pheromone, it is possible to achieve that each intelligent agent leaves information every time it moves, thereby improving the efficiency of the intelligent agent in finding the optimal path.

[0125] In one embodiment, Figure 6 As shown, before obtaining the object point heuristic value corresponding to each business object point, it also includes:

[0126] Step 602 : determining a first difference value between the current business object point and each next business object point; and determining a second difference value between each next business object point and the end business object point determination.

[0127] The first difference value may be a resource allocation account benefit between the current business object point and each next business object point.

[0128] The second difference value may be the resource allocation account income between each next business object point and the end business object point.

[0129] Specifically, the resource allocation account benefit between the point to be selected by the current agent and the point where it is located is determined, prompting the agent to select a point with higher benefits; the resource allocation account benefit between the target point and the candidate point is determined, prompting the agent to select a point closer to the target point.

[0130] Step 604: Determine the security factor corresponding to the agent based on the number of business object points in the business selection map domain space.

[0131] Among them, the safety factor can be to encourage the agent to choose a point that is feasible for itself.

[0132] Specifically, according to the number of business object points in the business selection map domain space, it is determined to encourage the agent to select feasible points for itself. m is the number of selectable points in the visual domain of the agent at the current point, and m′ is the number of unselectable points in the visual domain to meet the determined safety factors.

[0133] Step 606: Obtain a heuristic function corresponding to each business object point according to the first difference value, the second difference value, and the security factor.

[0134] Specifically, when each business object point has passability in any direction, the search strategy for each agent is determined. The heuristic function is essential in guiding the search process of each agent and plays an important role in the algorithm's rapid, effective, and accurate planning of feasible paths. When the agent moves from the current grid to the next grid, the selection probability of each candidate grid within the visible area can be calculated based on the heuristic function. The heuristic function can be represented by F(i, j, k), and its calculation process is as follows:

[0135]

[0136]

[0137]

[0138]

[0139] Where o is the current point, t is the candidate point, and d is the target point. M(i, j, k) is the resource allocation account benefit between the current agent's target point and the point it is currently located at, prompting the agent to choose a point with higher benefits. S(i, j, k) is a safety factor that encourages the agent to choose a feasible point. m is the number of selectable points in the agent's visual domain at the current point, and m′ is the number of unselectable points in the visual domain. N(i, j, k) is the resource allocation account benefit between the target point and the candidate point, prompting the agent to choose a point closer to the target point. w1, w2, and w3 are parameters representing the importance of each factor.

[0140] In this embodiment, the heuristic function corresponding to each business object point is established through the first difference value, the second difference value and the security factor, which can take into account all the influencing factors of calculating the heuristic value, so as to provide more accurate guidance when the intelligent agent uses the heuristic value to search, thereby improving the accuracy of the optimal path search.

[0141] In one embodiment, Figure 7 As shown, according to the first difference value, the second difference value and the security factor, the heuristic function corresponding to each business object point is obtained, including:

[0142] In step 702, the first difference value is used as the base and the first importance parameter is used as the power to obtain a first power function; and the second difference value is used as the base and the second importance parameter is used as the power to obtain a second power function; and the security factor is used as the base and the third importance parameter is used as the power to obtain a security power function.

[0143] The first importance parameter, the second importance parameter, and the third importance parameter may be weights corresponding to the first difference value, the second difference value, and the security factor, respectively.

[0144] Specifically, the first difference value M(i, j, k) is used as the base and the first importance parameter w1 is used as the power to obtain the first power function M(i, j, k) w1 ; and, using the second difference value N(i, j, k) as the base and the second importance parameter w3 as the power, obtain the second power function N(i, j, k) w3 ; and, taking the security factor S(i, j, k) as the base and the third importance parameter w2 as the power, we get the security power function S(i, j, k) w2 .

[0145] Step 704: Obtain the heuristic function corresponding to each business object point according to the first power function, the second power function, and the security power function.

[0146] Specifically, the heuristic function can be expressed as F(i, j, k), and its calculation process is as follows:

[0147]

[0148]

[0149]

[0150]

[0151] Where o is the current point, t is the candidate point, and d is the target point. M(i, j, k) is the resource allocation account benefit between the current agent's target point and the point it is currently located at, prompting the agent to choose a point with higher benefits. S(i, j, k) is a safety factor that encourages the agent to choose a feasible point. m is the number of selectable points in the agent's visual domain at the current point, and m′ is the number of unselectable points in the visual domain. N(i, j, k) is the resource allocation account benefit between the target point and the candidate point, prompting the agent to choose a point closer to the target point. w1, w2, and w3 are parameters representing the importance of each factor.

[0152] In this embodiment, by refining the process of constructing the heuristic function and introducing the first importance parameter, the second importance parameter and the third importance parameter, it is possible to provide different weights for the first difference value, the second difference value and the security factor, and modify the calculation accuracy of the heuristic function.

[0153] In one embodiment, Figure 8 As shown, after obtaining the business characteristic data corresponding to the resource allocation account set and establishing the business selection map domain space corresponding to the resource allocation account set based on the business characteristic data, it also includes:

[0154] Step 802 : Determine the position corresponding to the start business object point and the position corresponding to the end business object point in the business selection map domain space according to the business abstract environment model corresponding to the business selection map domain space.

[0155] The business abstract environment model may be an optimization model built based on business characteristic data corresponding to a resource allocation account set.

[0156] Specifically, based on the agent-related attributes in the customer information abstraction algorithm, and based on the map environment attributes in the resource allocation account information abstraction algorithm. The attributes of the resource allocation account include the resource allocation account type, installment payment, preferential activities, consumption limit, etc. Figure 9As shown, different points on the map represent resource allocation accounts with different attributes, randomly distributed across the map. The business selection map domain space, which serves as the product selection for the resource allocation account, determines the locations corresponding to the starting and ending business object points in the business selection map domain space based on the above information. Customer information includes whether the customer is new, basic information such as age, occupation, education level, income, and expenditure, available resource allocation account types and spending limits, key consumer expenditure characteristics, and social attribute information.

[0157] Step 804: Determine the starting movement direction of the agent according to the position corresponding to the starting business object point and the position corresponding to the ending business object point, and set the basic parameters of the agent algorithm and the global pheromone to be updated.

[0158] Specifically, based on the position corresponding to the starting business object point and the position corresponding to the ending business object point, as well as customer information including whether it is a new customer, age, occupation, education, income, expenditure and other basic information, the type of resource allocation account and consumption limit that can be handled, the main consumption expenditure characteristic information, social attribute information, etc., as parameters and constraints when the intelligent agent moves, the corresponding starting movement direction of the intelligent agent is determined, and the basic parameters of the intelligent agent algorithm and the global pheromone to be updated are set.

[0159] In this embodiment, by using the business abstract environment model to determine the corresponding positions of the starting business object point and the ending business object point, and further setting the basic parameters and initial global pheromones of the intelligent agent algorithm, the business abstract model can clear the parameter changes caused by operation and improve the accuracy of the model operation.

[0160] It should be understood that, although the steps in the flowcharts of the above embodiments are shown in sequence as indicated by the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be performed in other orders. Moreover, at least a portion of the steps in the flowcharts of the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily performed at the same time, but can be performed at different times. The execution order of these steps or stages is not necessarily to be performed in sequence, but can be performed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0161] Based on the same inventive concept, embodiments of the present application also provide a business object recommendation device for implementing the aforementioned business object recommendation method. The solution provided by this device is similar to the solution described in the aforementioned method. Therefore, the specific limitations in the one or more business object recommendation device embodiments provided below can be found in the above-mentioned limitations on the business object recommendation method and will not be further elaborated here.

[0162] In one embodiment, Figure 11 As shown, a business object recommendation device is provided, including: a map domain space construction module 1102, a candidate path construction module 1104, a target path determination module 1106 and a business recommendation module 1108, wherein:

[0163] The map domain space construction module 1102 is used to obtain business characteristic data corresponding to the resource allocation account set and establish a business selection map domain space corresponding to the resource allocation account set based on the business characteristic data; the business selection map domain space includes at least two business object points; the business object points are used to represent candidate business objects to be selected;

[0164] The candidate path construction module 1104 is configured to construct at least two candidate business object point selection paths based on each business object point. Each candidate business object point selection path has a corresponding selection probability. The selection probability is determined based on the directional probability between two adjacent business object points in the corresponding candidate business object point selection path. The directional probability is used to indicate the path formed by the agent passing through two adjacent business object points.

[0165] The target path determination module 1106 is configured to determine a target business object point selection path from among the candidate business object point selection paths; the target business object point selection path has a selection probability greater than the selection probabilities of the other business object point selection paths; the other business object point selection paths are candidate business object point selection paths other than the target business object point selection path from among the candidate business object point selection paths;

[0166] The business recommendation module 1108 is configured to recommend the target business object represented by the target business object point selection path to a target resource allocation account in the resource allocation account set.

[0167] In one embodiment, the target path determination module 1106 is further used to: when there is passability in any direction in each business object point, control the intelligent agent to perform a global search operation in the business selection map domain space, and update the global pheromone corresponding to the business selection map domain space; the global search operation is used to determine the operation of selecting the path of each candidate business object point in the business selection map domain space; when the number of times the global search operation is performed does not reach the preset number of iterations, return to the step of obtaining the business characteristic data corresponding to the resource allocation account set, and establish the business selection map domain space corresponding to the resource allocation account set based on the business characteristic data, until the number of global searches meets the preset number of iterations; when the number of times the global search operation is performed meets the preset number of iterations, determine the target business object point selection path in the business selection map domain space based on the updated global pheromone.

[0168] In one embodiment, the target path determination module 1106 is also used to: obtain the object point heuristic value corresponding to each business object point based on the existence of passability in any direction of each business object point; control the intelligent agent to move from the current business object point to the next business object point based on the object point heuristic value, and update the local pheromone corresponding to the current business object point; return to the execution of the global search operation in the business selection map domain space when passability exists in any direction of each business object point, and update the global pheromone corresponding to the business selection map domain space until each intelligent agent completes the search operation for the candidate business object point selection path; based on the search operation of each intelligent agent on the candidate business object point selection path, update the global pheromone corresponding to the business selection map domain space.

[0169] In one embodiment, the target path determination module 1106 is further used to: obtain the direction selection probability corresponding to each direction of the current business object point, and sort the direction selection probabilities; based on the sorting results corresponding to each direction selection probability, control the intelligent agent to select the direction selection probability to meet the preset probability conditions, move from the current business object point to the next business object point, and update the local pheromone corresponding to the business object point.

[0170] In one embodiment, the target path determination module 1106 is further used to: determine a first difference value between the current business object point and each next business object point; and determine a second difference value between each next business object point and the end business object point; determine a security factor corresponding to the intelligent agent based on the number of business object points in the business selection map domain space; and obtain a heuristic function corresponding to each business object point based on the first difference value, the second difference value and the security factor.

[0171] In one embodiment, the target path determination module 1106 is further used to: use the first difference value as the base and the first importance parameter as the power to obtain a first power function; and use the second difference value as the base and the second importance parameter as the power to obtain a second power function; and use the security factor as the base and the third importance parameter as the power to obtain a security power function; and obtain the heuristic function corresponding to each business object point based on the first power function, the second power function and the security power function.

[0172] In one embodiment, the map domain space construction module 1102 is also used to: determine the position corresponding to the starting business object point and the position corresponding to the ending business object point in the business selection map domain space based on the business abstract environment model corresponding to the business selection map domain space; determine the starting movement direction corresponding to the intelligent agent based on the position corresponding to the starting business object point and the position corresponding to the ending business object point, and set the basic parameters of the intelligent agent algorithm and the global pheromone to be updated.

[0173] Each module in the aforementioned business object recommendation device may be implemented in whole or in part through software, hardware, or a combination thereof. Each module may be embedded in or independent of a processor in a computer device in the form of hardware, or may be stored in a memory in the computer device in the form of software, so that the processor can call and execute the corresponding operations of each module.

[0174] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 12 As shown. The computer device includes a processor, a memory and a network interface connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store server data. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a business object recommendation method is implemented.

[0175] Those skilled in the art will understand that Figure 12 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0176] In one embodiment, a computer device is further provided, including a memory and a processor. The memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program.

[0177] In one embodiment, a computer-readable storage medium is provided, storing a computer program, which implements the steps in the above-mentioned method embodiments when executed by a processor.

[0178] In one embodiment, a computer program product or computer program is provided, the computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the steps of each of the above-described method embodiments.

[0179] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0180] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, database or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, distributed databases based on blockchains. The processor involved in the various embodiments provided herein may be, but are not limited to, a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic unit, a data processing logic unit based on quantum computing, and the like.

[0181] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0182] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A business object recommendation method, characterized in that: The method comprises: Obtaining business characteristic data corresponding to the resource allocation account set, and establishing a business selection map domain space corresponding to the resource allocation account set based on the business characteristic data; the business selection map domain space includes at least two business object points; the business object points are used to represent candidate business objects to be selected; At least two candidate business object point selection paths are constructed based on each of the business object points; each of the candidate business object point selection paths has a corresponding selection probability; the selection probability is determined based on a directional probability between two adjacent business object points in the corresponding candidate business object point selection path; the directional probability is used to indicate a path formed by an intelligent agent passing through two adjacent business object points; Determining a target business object point selection path from among the candidate business object point selection paths; wherein a selection probability corresponding to the target business object point selection path is greater than a selection probability corresponding to other business object point selection paths; wherein the other business object point selection paths are candidate business object point selection paths other than the target business object point selection path from among the candidate business object point selection paths; recommending the target business object represented by the target business object point selection path to a target resource allocation account in the resource allocation account set; After acquiring the business characteristic data corresponding to the resource allocation account set and establishing the business selection map domain space corresponding to the resource allocation account set based on the business characteristic data, the method further includes: Determining a position corresponding to a starting business object point and a position corresponding to an ending business object point in the business selection map domain space according to a business abstract environment model corresponding to the business selection map domain space; According to the position corresponding to the starting business object point and the position corresponding to the ending business object point, the starting movement direction corresponding to the agent is determined, and the basic parameters of the agent algorithm and the global pheromone to be updated are set.

2. The method according to claim 1, characterized in that Determining a target business object point selection path among the candidate business object point selection paths includes: When each of the business object points is passable in any direction, controlling the agent to perform a global search operation in the business selection map domain space and updating the global pheromone corresponding to the business selection map domain space; the global search operation is used to determine the selection path of each candidate business object point in the business selection map domain space; If the number of times the global search operation is performed does not reach the preset number of iterations, returning to the step of obtaining the business characteristic data corresponding to the resource allocation account set and establishing the business selection map domain space corresponding to the resource allocation account set based on the business characteristic data, until the number of times the global search operation is performed meets the preset number of iterations; When the number of times the global search operation is performed meets the preset number of iterations, the target business object point selection path is determined in the business selection map domain space according to the updated global pheromone.

3. The method according to claim 2, characterized in that When each of the business object points has passability in any direction, controlling the agent to perform a global search operation in the business selection map domain space and updating the global pheromone corresponding to the business selection map domain space includes: Based on the fact that each of the business object points has passability in any direction, obtaining an object point heuristic value corresponding to each of the business object points; Based on the object point heuristic value, controlling the agent to move from the current business object point to the next business object point, and updating the local pheromone corresponding to the current business object point; In the case that any of the agents fails to complete the search operation for the candidate business object point selection path, returning to the case where there is passability in any direction at each of the business object points, controlling the agents to perform a global search operation in the business selection map domain space, and updating the global pheromone corresponding to the business selection map domain space until each of the agents completes the search operation for the candidate business object point selection path once; Based on the search operation of each agent on the candidate business object point selection path, the global pheromone corresponding to the business selection map domain space is updated.

4. The method according to claim 3, characterized in that The controlling the agent to move from the current business object point to the next business object point and updating the local pheromone corresponding to the current business object point includes: Obtaining direction selection probabilities corresponding to various directions of the current business object point, and sorting the direction selection probabilities; Based on the sorting results corresponding to each of the direction selection probabilities, the agent is controlled to select the direction selection probability that meets the preset probability condition, move from the current business object point to the next business object point, and update the local pheromone corresponding to the business object point.

5. The method according to claim 3, characterized in that Before obtaining the object point heuristic value corresponding to each of the business object points, the method further includes: Determining a first difference value between the current business object point and each of the next business object points; and determining a second difference value between each of the next business object points and an end business object point determination; Determining a security factor corresponding to the agent according to the number of each of the business object points in the business selection map domain space; A heuristic function corresponding to each of the business object points is obtained according to the first difference value, the second difference value, and the security factor.

6. The method according to claim 5, characterized in that The obtaining, according to the first difference value, the second difference value, and the security factor, a heuristic function corresponding to each of the business object points includes: Using the first difference value as a base and the first importance parameter as a power to obtain a first power function; and using the second difference value as a base and the second importance parameter as a power to obtain a second power function; and using the security factor as a base and the third importance parameter as a power to obtain a security power function; According to the first power function, the second power function and the security power function, a heuristic function corresponding to each of the business object points is obtained.

7. A business object recommendation device, characterized in that: The device comprises: A map domain space construction module is configured to obtain business characteristic data corresponding to a resource allocation account set and, based on the business characteristic data, establish a business selection map domain space corresponding to the resource allocation account set; the business selection map domain space includes at least two business object points; the business object points are used to represent candidate business objects to be selected; A candidate path construction module is configured to construct at least two candidate business object point selection paths based on each of the business object points; each candidate business object point selection path has a corresponding selection probability; the selection probability is determined based on a directional probability between two adjacent business object points in the corresponding candidate business object point selection path; the directional probability is used to indicate a path formed by an intelligent agent passing through two adjacent business object points; a target path determining module, configured to determine a target business object point selection path from among the candidate business object point selection paths; wherein the selection probability corresponding to the target business object point selection path is greater than the selection probabilities corresponding to other business object point selection paths; and wherein the other business object point selection paths are candidate business object point selection paths other than the target business object point selection path from among the candidate business object point selection paths; A business recommendation module, configured to recommend the target business object represented by the target business object point selection path to a target resource allocation account in the resource allocation account set; The map domain space construction module is also used to: Determining a position corresponding to a starting business object point and a position corresponding to an ending business object point in the business selection map domain space according to a business abstract environment model corresponding to the business selection map domain space; According to the position corresponding to the starting business object point and the position corresponding to the ending business object point, the starting movement direction corresponding to the agent is determined, and the basic parameters of the agent algorithm and the global pheromone to be updated are set.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images