Differentiated resource allocation method and device based on user requirements

Abstract

The embodiment of the invention discloses a differentiated resource allocation method and device based on user requirements. The method comprises the steps that a first configuration combination is generated based on a user intention and an object configuration strategy, and the first configuration combination comprises a plurality of different types of objects conforming to the user intention and first configuration weights corresponding to the objects; the method includes generating a plurality of second configuration combinations generalizedly deviating from the first configuration combination based on the adjacent dispersion, the second configuration combinations including a plurality of different types of objects conforming to the user's intent and second configuration weights corresponding to the objects. The method includes obtaining a plurality of second configuration combinations related to a user input based on the user input. The method further comprises the step of selecting an optimal second configuration combination to be output to the user. In this way, a plurality of second configuration combinations are generalized on the basis of the first configuration combination, so that the objects in the combinations have significant differences, and resource configuration can meet the differentiated requirements of users.

Description

Technical Field

[0001] One or more embodiments of this specification relate to the field of computer technology, and particularly to a method and device for differential resource allocation based on user requirements. Background Art

[0002] With the development of Internet technology, more and more data is transmitted or stored through the Internet. With the growth of the amount of data in the Internet, a variety of data-based services have emerged, and the industries involved in these services are relatively wide-ranging.

[0003] When a user requests a service, the resource allocation generated and returned by the system for the service is the same for all users, which cannot meet the specific intentions of users. Summary of the Invention

[0004] One or more embodiments of this specification describe a method and device for differential resource allocation based on user requirements.

[0005] In the first aspect of this specification, a method for differential resource allocation based on user requirements is provided. The method includes generating a first configuration combination based on a configuration policy of user intent and objects, where the first configuration combination includes multiple different types of objects that meet the user intent and a first configuration weight corresponding to the objects. The method further includes generating a plurality of second configuration combinations that are generally deviated from the first configuration combination based on adjacent dispersion, where the second configuration combinations include multiple different types of objects that meet the user intent and a second configuration weight corresponding to the objects. The method further includes obtaining a plurality of second configuration combinations related to the user input based on the user input. In addition, the method further includes selecting an optimal second configuration combination from them and outputting it to the user.

[0006] In the second aspect of this specification, a device for differential resource allocation based on user requirements is provided. The device includes a first generation module configured to generate a first configuration combination based on a configuration policy of user intent and objects, where the first configuration combination includes multiple different types of objects that meet the user intent and a first configuration weight corresponding to the objects. The device further includes a second generation module configured to generate a plurality of second configuration combinations that are generally deviated from the first configuration combination based on the first configuration combination and adjacent dispersion, where the second configuration combinations include multiple different types of objects that meet the user intent and a second configuration weight corresponding to the objects. In addition, the device further includes an output module configured to obtain a plurality of second configuration combinations related to the user input based on the user input, and select an optimal second configuration combination from them and output it to the user.

[0007] In a third aspect of this specification, an electronic device is provided. The electronic device includes a processor and a memory, and the processor is connected to the memory. The memory is used to store executable program code. The processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to execute the above method.

[0008] In a fourth aspect of this specification, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the above method is implemented.

[0009] It should be understood that the content described in the Summary of the Invention section is not intended to limit the key or important features of the embodiments of the present disclosure, nor is it used to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] To more clearly illustrate the technical solutions in the embodiments of this specification, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings in the following description are only some embodiments of this specification. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0011] Figure 1 A schematic diagram showing an example environment in which multiple embodiments of this specification can be implemented;

[0012] Figure 2 A flowchart showing a method for differential resource allocation based on user requirements in some embodiments of this specification;

[0013] Figure 3 A schematic diagram showing an example system architecture for implementing differential resource allocation based on user requirements in some embodiments of this specification;

[0014] Figure 4 A schematic flowchart showing a process for generating a first configuration combination in some embodiments of this specification;

[0015] Figure 5 A schematic flowchart showing a process for configuring a first configuration combination in some embodiments of this specification;

[0016] Figure 6 A schematic flowchart showing a judgment process for configuring a first configuration combination in some embodiments of this specification;

[0017] Figure 7 A schematic diagram showing a principle process for generating a second configuration combination in some embodiments of this specification;

[0018] Figure 8A flowchart showing the process of generating a second configuration combination for some embodiments of this specification;

[0019] Figure 9 A schematic diagram showing an example process applied to a human - machine interaction scenario for some embodiments of this specification;

[0020] Figure 10 A schematic block diagram showing a differential resource configuration device based on user requirements for some embodiments of this specification;

[0021] Figure 11 A schematic diagram of the structure of an electronic device for some embodiments of this specification. Detailed implementation manners

[0022] Next, the technical solutions in the embodiments of this specification will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of this specification.

[0023] Terms such as "first", "second", "third", etc. in the specification, claims and the above - mentioned drawings of this specification are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "include" and "have" and any variations thereof are intended to cover non - exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally further includes steps or units not listed, or optionally further includes other steps or units inherent to these processes, methods, products or devices.

[0024] As described above, the system responds to a service request and generates the same resource configuration for the same type of services according to the standard configuration criteria. Since the service execution result is affected by the resource configuration, when each object within the service is uniformly configured according to the service type, when the same - type services are executed according to the same resource configuration, the same result will be produced. Not only is the resource configuration based on the service one - size - fits - fits - all, but also the result of executing the service is one - size - fits - fits - all, which cannot meet the personalized needs of users.

[0025] In some related technologies, although user requirements are considered and each object within the service performs different resource configurations according to the service type and user requirements, the differences between different services are not obvious, only the basic needs of users are matched. For users with similar needs, there are still differences in their user intents, and the resource configurations that meet the basic needs cannot match multiple differential user intents. In addition, the service execution result also cannot effectively meet the user intent.

[0026] To this end, the embodiments of this specification propose a differential resource allocation method based on user needs. In the embodiments of this specification, based on the user needs and the configuration strategy of the object, a first configuration combination that is basically matched with the user intention is generated. Taking the first configuration combination as an anchor, through the generalization deviation method, a plurality of second configuration combinations with differential features are obtained to further match the user intention. The first configuration combination and the second configuration combination contain the resource allocation information for each object.

[0027] In this way, the second configuration combinations generated based on the first configuration combination can further capture the personalized needs of users. Even if users have similar needs, the generated multiple configuration combinations can comprehensively match the differential intentions among different similar users, achieving personalized experiences for each user. And the business execution results based on the second configuration combinations can match the user intentions to achieve personalized execution results.

[0028] Figure 1 FIG. 100 shows an example environment in which multiple embodiments of this specification can be implemented. As Figure 1 shown, the environment 100 includes a computing unit 110, and the computing unit 110 can be any device with computing power or processing power. For example, the computing unit 110 can include, but is not limited to, mobile phones, tablets, desktop computers, servers, etc. The differential resource allocation method based on user needs can be executed in the computing unit 110. As Figure 1 shown, the user intention 102 and the configuration strategy 104 of the object are included in the environment 100. The user intention 102 can be information reflecting the user needs, including object preferences and user types. That is to say, the generated configuration combination needs to conform to the user's preferences for the objects involved in the business and be configured according to the user type. The configuration strategy 104 of the object can be the strategy on how the object is configured in the business. For example, different strategies can be configured according to the type of the object.

[0029] As Figure 1 shown, the computing unit 110 can generate a first configuration combination 106 based on the user intention 102 and the configuration strategy 104 of the object. The first configuration combination includes multiple different types of objects that conform to the user intention and the first configuration weights corresponding to the objects. The generated first configuration combination can match the user needs. In some recommendation scenarios, the corresponding first configuration weights can be configured according to the recommended objects so that the generated first configuration combination can match the user needs.

[0030] As Figure 1As shown, the computing unit 110 can also generate multiple second configuration combinations 108 in a generalized offset manner based on the first configuration combination 106. The second configuration combination 108 includes multiple different types of objects that meet the user's intention and corresponding second configuration weights. The second configuration weights are obtained by offsetting based on the first configuration weights. After each offset, the resulting second configuration combination has a certain degree of discreteness compared to the previous configuration combination (i.e., the adjacent discreteness referred to in this article, which is used to characterize that there is a certain difference between two configuration combinations). In this way, the generated second configuration combinations have differentiated characteristics, that is, the configuration weights of different types of objects are different, and can better meet the user's personalized needs.

[0031] In some recommendation scenarios, based on the first configuration combination generated by the recommendation service, multiple second configuration combinations are obtained through generalized deviation, and there is a certain degree of discreteness between each generated second configuration combination and the adjacent configuration combination. In this way, the multiple second configuration combinations are equivalent to being generated based on different degrees of object preferences and user types, and different second configuration weights are configured for different types of recommended objects, which can more comprehensively match the user's personalized needs. The above-mentioned recommendation scenarios can include scenarios such as investment business recommendations and design business recommendations. The design business can include business such as information center energy portfolio design and production design.

[0032] It should be understood that the architecture and functions in the example environment 100 are described only for exemplary purposes, and do not imply any limitation on the scope of the present disclosure. Embodiments of the present disclosure can also be applied to other environments with different structures and / or functions.

[0033] Figure 2 The flowchart of the differential resource allocation method 200 based on user requirements according to some embodiments of the present specification is shown. The method 200 can be executed, for example, by a server, or by a terminal device, or by the server and the terminal device cooperating with each other. As Figure 2 shown, at block 202, the method 200 can generate a first configuration combination based on the user's intention and the configuration strategy of the object, where the first configuration combination includes multiple different types of objects that meet the user's intention and corresponding first configuration weights. In some embodiments, based on the object preferences in the user's intention, objects related to the object preferences are obtained. For example, if the object preferences are objects of type A and type C, then objects belonging to type A and type C are obtained. And, based on the user type in the user's intention, the configuration strategy under this user type is obtained, and the configuration strategy related to the preferred object is obtained therefrom. Then, based on the configuration strategy corresponding to the corresponding object type under the corresponding user type, the first configuration weights are respectively configured for the objects of type A and type C to form the first configuration combination.

[0034] In block 204, method 200 may generate a plurality of second configuration combinations that deviate from the generalization of the first configuration combination based on the adjacent divergence, where the second configuration combinations include a plurality of different types of objects that meet the user's intention and second configuration weights corresponding to the objects. The adjacent divergence is obtained by calculating the divergence value between two configuration combinations, and is used to characterize the degree of difference between the configuration weights of each object within the two configuration combinations during adjacent offset times. In this specification, by aiming at a certain divergence between two configuration combinations during adjacent offset times, the second configuration combination after each offset is calculated. In some embodiments, starting from the first configuration combination, a plurality of second configuration combinations are obtained through multiple offsets. Each offset obtains a second configuration combination by offsetting a certain offset amount relative to the first configuration combination, and there is a certain divergence between the second configuration combination and the adjacent configuration combination. For example, during the first offset, there is a certain divergence between the second configuration combination obtained by the first offset and the first configuration combination. During the second offset, there is a certain divergence between the second configuration combination obtained by the second offset and the second configuration combination obtained by the first offset.

[0035] In the above manner, based on the user's intention and the configuration strategy of the object, the first configuration combination is generated from the level of combining qualitative and quantitative aspects, which can initially match the user's needs. Further, taking the first configuration combination as an anchor, a plurality of second configuration combinations are generated through the generalization offset method. The plurality of second configuration combinations are generated based on the different degrees of object preference and user type, and can more comprehensively match the user's personalized needs.

[0036] The above describes specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve the desired results. Additionally, the processes depicted in the figures do not necessarily require the particular order or sequential order shown to achieve the desired results. In certain implementations, multitasking and parallel processing are also possible or may be advantageous.

[0037] Figure 3 A schematic diagram of an example system architecture 300 for implementing differential resource allocation based on user needs according to some embodiments of this specification is shown. As Figure 3As shown, the system architecture 300 includes an object pool 302 and a policy pool 306. The object pool 302 stores multiple objects, including different types of objects and multiple objects of each type. The policy pool 306 stores configuration policies corresponding to different types of objects (the configuration policy for each type of object is the sub-configuration policy referred to in the text). The object pool 302 and the policy pool can be updated regularly to ensure that resource configuration is generated based on updated policy criteria and updated objects, better meeting the user needs that change with the increasingly changing data. The system architecture 300 also includes a user profile 308, which is generated based on the user's historical behavior, such as which objects the user has historically selected and the user types that the object can provide. The data involved in this specification (including user profiles, personal information, etc.) are all information and data that have been authorized by the user or fully authorized by all parties, and the collection of relevant data complies with the relevant laws, regulations and standards of the relevant countries and regions.

[0038] As Figure 3As shown, the system architecture 300 further includes a first generation module 310 that generates a first configuration combination 312 based on a user intention and a configuration policy of an object. The user intention is determined based on a user profile 306, and an object preference and a user type are extracted from the user's historical profile using a semantic recognition model. After determining the user intention, a combination 304 that matches the object preference in the user intention can be obtained from an object pool. In some embodiments, the combination 304 includes combinations composed of all types of objects in the object pool 302, and the combinations can be divided into several categories, such as Category I combinations, Category II combinations, etc., and the type combinations of the objects included in each category of combinations are different. Taking the application scenario of the investment business as an example, different combinations can be formed according to different types of product combinations. For example, Category I combinations include public offering fund products of multiple types (such as stocks, futures, bonds), Category II combinations include private placement fund products of multiple types (such as stocks, futures, bonds), Category III combinations include public offering fund products and insurance-type fund products, and Category IV combinations include private placement fund products and insurance-type fund products. Taking the application scenario of the energy combination design of the information center as an example, different combinations can be formed according to different types of power generation equipment combinations. For example, Category I combinations include thermal power generation equipment and wind power generation equipment, Category II combinations include thermal power generation equipment and solar power generation equipment, Category III combinations include thermal power generation equipment and tidal energy power generation equipment, Category IV combinations include thermal power generation equipment, wind power generation equipment, and solar power generation equipment, Category V combinations include thermal power generation equipment, wind power generation equipment, and tidal energy power generation equipment, and Category VI combinations include thermal power generation equipment, wind power generation equipment, solar power generation equipment, and tidal energy power generation equipment. The first generation module 310 obtains a certain category of combination from the combination 304 based on the object preference. Then, the first generation module 310 obtains the configuration policy of the object of this type in the policy pool according to the object types in the obtained certain category of combination. Finally, the first generation module performs a first configuration weight configuration on this category of combination according to the obtained configuration policy, that is, generates the first configuration combination 312. In an example, when a Category I combination is obtained from the combination 304 according to the user intention, and there are multiple Category I combinations. Correspondingly, the configuration policies corresponding to the object types in the Category I combination are also obtained from the policy pool, and then multiple first configuration combinations are generated, and each first configuration combination corresponds to a Category I combination. When there are multiple user intentions, the first generation module can generate corresponding first configuration combinations according to each type of combination in the combination 304.

[0039] Such as Figure 3As shown, the system architecture 300 further includes a second generation module 314, which is configured with a generalization model that generates a plurality of second configuration combinations 316 based on generalization of a first configuration combination 312. When there are multiple types of first configuration combinations and there are multiple first configuration combinations in each type of first configuration combination, the second generation module 314 can generate a plurality of second configuration combinations based on generalization of each first configuration combination. In some embodiments, there are 1 to N first configuration combinations of type I, and each first configuration combination can generate M second configuration combinations through generalization, where N and M are natural numbers. Thus, a total of N * M second configuration combinations are generated through generalization of N first configuration combinations of type I. Correspondingly, other types of first configuration combinations are also generalized to generate a plurality of second configuration combinations in the same way.

[0040] Figure 4 FIG. shows a schematic diagram of a process 400 for generating a first configuration combination according to some embodiments of the present specification. For ease of understanding, a single combination is used as an example for illustration. As Figure 4 shown, a formation unit 406 of the first generation module 310 determines a certain combination that conforms to a user intention 402 from a combination 404 based on the user intention 402, that is, obtains a first configuration combination 408 without configured first configuration weights, such as combination 1 of type I as the first configuration combination 1 of type I. A configuration unit 412 of the first generation module 310 configures the first configuration combination 408 without configured first configuration weights based on a configuration policy 410 related to the object type in the combination, and finally obtains a first configuration combination 414 configured according to the configuration policy 410. For example, when combination 1 of type I includes object A, object C, and object E, then a configuration policy for object A, a configuration policy for object C, and a configuration policy for object E are obtained from an object pool. After that, the configured first configuration combination of type I includes object A and its configuration weight W1, object C and its configuration weight W3, and object E and its configuration weight W5.

[0041] The configuration unit 412 can determine the first configuration weights of objects of each type based on the configuration policy. In some embodiments, objects in the first configuration combination can be quantitatively configured according to the configuration policy of each type. In some embodiments, the configuration policy of each type can be constructed into a constraint condition, and based on a target model, a dynamically optimized first configuration weight is calculated. In some examples, with the goal of maximizing the total score of the first configuration weights in the first configuration combination, for example, if the first configuration combination includes object A, object C, and object E, then Score 总 = Score W1 + Score W3 + Score W5 , and the objective function is Max(Score 总) Among them, the scoring value of the first configuration weight corresponding to each type of object is determined based on the scoring criteria under different application scenarios. This scoring value can be based on the objective score obtained from model operations or can be determined based on empirical values. Correspondingly, for the first configuration weights of different types of objects in the first configuration combination, multiple constraint conditions can be established according to the sub-configuration strategies corresponding to the object types. For example, the deviation between the first configuration weight and the reference configuration weight is within the deviation range. Among them, the reference configuration weight can be determined based on empirical values. Through the above constraint conditions and the objective function, the first configuration weights of each object in the first configuration combination can be optimized and calculated.

[0042] Through the above method, it is possible to generate the first configuration combination that meets the user's needs according to the user's intention and the configuration strategy of the object, and multiple first configuration combinations can be generated according to multiple user intentions, which can match the personalized needs of different users.

[0043] Figure 5 The figure shows a schematic diagram of a process 500 for configuring the first configuration combination in some embodiments of the present specification. This process 500 configures the first configuration combination 502 twice. The first time is to perform a preliminary configuration on the first configuration combination, and the second time is to perform a re-configuration on the first configuration combination, and a floating adjustment is made based on the configuration weight of the preliminary configuration as a reference. In some scenarios, the information center energy portfolio design is preliminarily configured based on the environmental protection level or energy-saving level of power generation. The proportion of the number of thermal power generation equipment units in the total number of power generation equipment units can be determined according to the environmental protection level or energy-saving level, or the proportion of the power generation power of thermal power generation equipment in the total power generation power of all discharge equipment can be determined. Correspondingly, the total configuration weight of other power generation equipment can also be determined. In some scenarios, the investment business is preliminarily configured based on the risk level. The investment proportion of equity fund products can be determined according to the risk level. Correspondingly, the total configuration weight of other investment products can also be determined.

[0044] Such as Figure 5, the first configuration subunit 506 performs preliminary configuration on the first configuration combination 502 without the first configuration weight based on the first policy in the policy pool 504. The first policy can be a policy set for a specific benchmark, such as a risk level, an environmental protection level, or an energy-saving level, and is used to configure the first configuration weight for a certain type of object and other types of objects. In some examples, the first policy determines that the proportion of the first configuration weight of type A objects in the total configuration weight (i.e., the configuration proportion of type A objects) T1 = W1 / (W1 + W3 + W5), and determines that the sum of the first configuration weights of other types of objects except type A objects is (1 - T1) (see box 508). The second configuration subunit 510 performs re-configuration based on the second policy and the result of the preliminary configuration by the first configuration subunit 510. The second policy can be a policy for optimizing the first configuration weight of each type of object. An optimization model is configured in the second configuration subunit 510, and optimization calculations can be performed according to the objective function and constraint conditions. Among them, the constraint conditions not only need to establish multiple constraint conditions corresponding to the first configuration weights of different types of objects according to the sub-configuration policies corresponding to the object types, but also need to float within a certain range based on the result of the preliminary configuration by the first configuration subunit as the benchmark. The first configuration weights of each type of object configured by the second configuration subunit need to be within the above range. In some examples, taking the maximization of the total score of the first configuration weights in the first configuration combination as the objective function Max(Score 总 ), W1, W3, and W5 each determine their constraint conditions according to the sub-configuration policies. In addition, W1 also needs to be within the range of (T1 - △T, T1 + △T), and (W3 + W5) also needs to be within the range of (1 - T1 - △T’, 1 - T1 + △T’). Among them, △T and △T’ are the configuration weight values allowed for reasonable floating, such as floating up and down by 5 percentage points.

[0045] In this way, when the object preference is concretized to a specific benchmark (such as a risk level or an energy-saving level, etc.), the first configuration combination composed of various types of objects under the specific benchmark can be matched according to the user's intention, which is closer to the user's differentiated resource configuration requirements.

[0046] Figure 6 The figure shows a schematic diagram of the judgment process 600 for configuring the first configuration combination in some embodiments of this specification. When the second configuration subunit 510 performs re-configuration, it needs to make a judgment to determine whether a certain type of object that needs to be quantitatively configured is included in the combination. If so, it executes the left process; if not, it executes the right process. In some embodiments, box 602 determines whether a second type of object is included in the combination configured by the first configuration subunit 506. If so, based on the sub-configuration policy of the second type of object, the first configuration weight of the second type of object is quantitatively configured 604, and Figure 5In the example, the optimization model optimizes and calculates the first configuration weights of other types of objects except the second type of objects. The sum of the first configuration weights of all objects calculated in block 604 and block 606 is equal to 1. The first configuration weight of the first type of objects calculated in the optimization model needs to fluctuate within a certain range based on the result T preliminarily configured by the first configuration subunit. The sum of the first configuration weights of other types of objects except the first type of objects calculated in the optimization model and the first configuration weight of the second type of objects obtained in block 604 needs to fluctuate within a certain range based on the result (1 - T) preliminarily configured by the first configuration subunit. If the judgment result of block 602 is no, the optimization model is used for calculation. In some examples, the second type refers to type B. When there is a type B object in the combination, the calculation is performed according to the left process, otherwise according to the right process. Taking the investment business scenario as an example, the type A objects in the figure are equity products, the type B objects are insurance products, and the type C and type E objects are money market and bond products respectively. In the second strategy, quantitative allocation is performed on insurance fund products, and the allocation of other types (equity, money market, bond) of products is generated using the optimization model. The sum of the first configuration weights of all types of objects generated, W1 + W2 + W3 + W4 = 1, and W1 meets the constraint conditions specified by its sub-allocation strategy and W1 is within the range of (T1 - △T, T1 + △T), and W2 meets the constraint conditions specified by its sub-allocation strategy, and (W2 + W3 + W4) is within the range of (1 - T1 - △T’, 1 - T1 + △T’). Taking the energy portfolio design of the information center as an example, the utilization rate of tidal power generation equipment is low. In the second strategy, quantitative allocation is performed on this type of power generation equipment. When the type B object is a tidal power generation equipment, the number or power generation capacity of the tidal power generation equipment is configured according to the quantitative allocation. Other power generation equipment is generated using the optimization model.

[0047] In this way, the first configuration weights of each object in the first configuration portfolio can be configured together according to the quantitative allocation strategy and the dynamic allocation strategy. The first configuration portfolio required for this scenario can be generated according to the allocation strategy under different scenario requirements. The method of the embodiments of this specification can be flexibly configured according to the characteristics of the scenario, and a resource allocation portfolio that meets the user's needs can be generated based on various complex or simple allocation strategies.

[0048] Figure 7 The schematic diagram of the principle process 700 for generating the second configuration portfolio according to some embodiments of this specification is shown. As Figure 7As shown, multiple second configuration combinations are obtained by generalizing and offsetting based on the first configuration combination. In the figure, each second configuration combination is offset and controlled relative to the first configuration combination as the starting point, and a certain adjacent discreteness needs to be maintained between adjacent combinations during the offset. For example, the offset is generalized with the maximum discreteness as the goal. In some embodiments, starting from the first configuration combination, the first offset amount is offset to generate multiple candidate second configuration combinations. With the goal of maximizing the adjacent discreteness between the first second configuration combination and the first configuration combination, one second configuration combination is determined from the multiple candidate second configuration combinations, that is, the second configuration combination 1 shown in the figure. Then, starting from the first configuration combination, the second offset amount is offset to generate multiple candidate second configuration combinations. With the goal of maximizing the adjacent discreteness between the second second configuration combination and the first second configuration combination, one second configuration combination is determined from the multiple candidate second configuration combinations, that is, the second configuration combination 2 shown in the figure. Then, and so on, offset m times to obtain m second configuration combinations. The circles containing the first configuration combination and the second configuration combinations obtained by each offset are the feasible solution spaces for solving the second configuration combinations during each offset. This feasible solution space can be determined by setting the first constraint condition that the current offset amount does not exceed the offset threshold. In some embodiments, with the goal of maximizing the adjacent discreteness between the second configuration combination under the current offset and the combination under the previous offset, and the current offset amount satisfies the first constraint condition that does not exceed the offset threshold, the second configuration combination under the current offset is determined. Among them, the combination under the previous offset can be the second configuration combination determined under the previous offset, or the first configuration combination as the starting point can be regarded as the configuration combination under the 0th offset.

[0049] In some scenarios, other constraint conditions also need to be considered in the feasible solution space. For example, the constraint conditions of the influence factors of the object in different dimensions. For example, investment products are affected by various risk factors, such as risk factors such as industry exposure. Then, a second constraint condition needs to be set for the second configuration weight of the investment product affected by it, which can be that the second configuration combination and the first configuration combination satisfy the second constraint condition of the same industry exposure. Another example is that power generation equipment is affected by various environmental factors. Then, a second constraint condition needs to be set for the second configuration weight of the power generation equipment affected by it, which can be that the second configuration combination and the first configuration combination satisfy the second constraint condition of the same environmental impact.

[0050] Figure 8 Shows the process 800 for generating the second configuration combination in some embodiments of this specification. As Figure 8, the offset unit 804 of the second generation module 314 performs offset processing based on the first matrix 802 to obtain the candidate second matrices 806-11, …, 806-mk after each offset, where m and k are natural numbers. The first matrix 802 is formed based on the first configuration combination and is a matrix composed of an object and its first configuration weights. The offset unit 804 performs m deviations based on Figure 7 the offset principle, and each offset consists of k candidate second matrices. Among them, the offset matrix is composed of an object and its offset amount. The candidate second matrix is obtained by superimposing the offset matrix on the first matrix. Then, the generation unit 810 of the second generation module 314 generalizes and generates m second matrices 812-12, …, 812-m1 based on the candidate second matrices using the generalization model 808. The generalization model 808 determines the second configuration combination under the current offset according to the maximum adjacent discreteness objective function and the offset amount satisfying the first constraint condition. In some embodiments, the objective function is to maximize the adjacent discreteness between the second matrix under the current offset and the first matrix or the second matrix under the previous offset, and the offset amount in the offset matrix under the current offset satisfies the first constraint condition of not exceeding the offset threshold, and the offset matrix under the current offset is calculated. After calculating the offset matrix, the second matrix can be obtained by superimposing the offset matrix on the first matrix. In some examples, the generalization model is constructed in matrix form and optimized operations are performed. The generation unit determines the optimal candidate second matrix 806-11 from the candidate second matrices of the first offset as the second matrix 812-12, and the second matrix 812-12 represents the second configuration combination obtained from the first offset. Each offset can obtain the corresponding optimal candidate second matrix and determine it as the second matrix corresponding to the second configuration combination. At the m-th offset, the optimal candidate second matrix 806-m1 is determined from the candidate second matrices of the m-th offset as the second matrix 812-m1.

[0051] In addition, when other constraint conditions need to be considered in some scenarios, the generalization model also needs to consider other constraint conditions. In some embodiments, the objective function is to maximize the adjacent discreteness between the second matrix under the current offset and the first matrix or the second matrix under the previous offset, and the offset amount in the offset matrix under the current offset satisfies the first constraint condition of not exceeding the offset threshold, and the second matrix under the current offset and the first matrix satisfy the second constraint condition of the same industry exposure, and the offset matrix under the current offset is calculated. Then, based on the first matrix and the offset matrix under the current offset, the second matrix under the current offset is determined.

[0052] In the above manner, aiming at the maximum adjacent discreteness between adjacent combinations and ensuring a certain offset between the second configuration combination and the first configuration combination and meeting the constraint conditions, a plurality of second configuration combinations are generally generated. The second configuration combination is generated based on the first configuration combination, which can not only match the user's needs, but also the objects within the generated second configuration combination have significant difference characteristics, and the business results executed based on this resource configuration can better meet the user's expectations. For example, according to the second investment configuration combination obtained by generalization, when purchasing investment products, the rate of return for a certain period is close to the rate of return expected by the user. Another example is that according to the second energy configuration combination obtained by generalization, an energy system is constructed and power is generated, and the power generation capacity basically reaches the power generation capacity expected by the user.

[0053] Figure 9 The schematic diagram shows an example process 900 applied to a human-computer interaction scenario in some embodiments of this specification. The user 902 sends a service request to the robot 904 on the client (computer device). The robot 904 responds to the service request, determines the second configuration combination that best matches the user's intention, and then the robot 904 will display it on the interface of the client and return it to the user. Among them, the computer device can be, but is not limited to, various personal computers, laptop computers, smart phones, tablet computers, and portable wearable devices.

[0054] Such as Figure 9 , the output module 908 can be set on the computing unit 110 shown in Figure 1 or on other computing units independent of the computing unit 110 shown in Figure 1 . The computing unit can be a computing unit on the client or a computing unit on the server. A plurality of second configuration combinations are stored in the combination pool 906. The second configuration combinations are pre-generated according to the user intentions of different users and the configuration strategies of the objects, based on the method for differentiating resource configuration based on user needs provided by the embodiments of this specification. When a user sends a user input to the robot 904, the user input is semantically recognized to determine the user intention, that is, to determine which types of objects the required second configuration combination consists of. The output module 908 selects a plurality of second configuration combinations that meet the user intention from the combination pool 906. A matching model is configured in the output module 908 to calculate the matching degree between the second configuration combination and the user input. In one example, a weight score table is configured in the output module 908, and the weight scores of different types of objects are stored in the weight score table. The scores of a plurality of second configuration combinations are calculated respectively. For example, using the second configuration weights and weight scores in the second configuration combination, the weighted sum is obtained as the matching score of the second configuration combination. The second configuration combination with the maximum or minimum matching score is returned to the user as the optimal combination. In addition, the output module 908 can also obtain the optimal second configuration combination based on other quantization methods. After that, the user determines whether to execute according to the second configuration combination.

[0055] In this way, the robot can quickly respond to user input, obtain a second configuration combination that matches it from the combination pool, and return the optimal configuration combination to the user. The processing efficiency is high, and the latest configuration combination that meets the user's personalized needs can be fed back based on the updated combination pool.

[0056] Figure 10 FIG. shows an exemplary block diagram of a differential resource allocation device 1000 based on user requirements according to some embodiments of the present specification. The device 1000 includes a first generation module 1002 and a second generation module 1004.

[0057] The first generation module 1002 is configured to generate a first configuration combination based on the user intention and the configuration strategy of the object. The first configuration combination includes multiple different types of objects that meet the user intention and a first configuration weight corresponding to the object.

[0058] The second generation module 1004 is configured to generate a plurality of second configuration combinations that are generally deviated from the first configuration combination based on the adjacent discreteness based on the first configuration combination. The second configuration combination includes multiple different types of objects that meet the user intention and a second configuration weight corresponding to the object.

[0059] In some embodiments, the first generation module 1002 includes a formation unit, a configuration unit, and an intention extraction unit. The formation unit is configured to form a first configuration combination composed of multiple different types of objects that meet the user intention based on the objects obtained from the object pool that meet the user intention. The configuration unit is configured to configure a first configuration weight for the objects in the first configuration combination based on the configuration strategy. The intention extraction unit is configured to determine the user intention based on the user's historical profile.

[0060] In some embodiments, the second generation module 1004 includes an offset unit and a generation unit. The offset unit is configured to start from the first configuration combination and obtain a plurality of second configuration combinations through multiple offsets. The generation unit is configured to, for each offset, target to maximize the adjacent discreteness between the second configuration combination under the current offset and the configuration combination under the previous offset, and determine the second configuration combination under the current offset when the current offset amount satisfies a first constraint condition that does not exceed the offset threshold.

[0061] In some embodiments, the device 1000 further includes an output module, which is configured to obtain a plurality of second configuration combinations related to the user input based on the user input, and select an optimal second configuration combination therefrom to output to the user.

[0062] As Figure 11, an electronic device 1100 provided by an embodiment of this specification may include: at least one processor 1102, at least one network interface 1108, a user interface 1106, a memory 1110, and at least one communication bus 1104.

[0063] Among them, the communication bus can be used to realize the connection and communication of the above-mentioned various components.

[0064] Among them, the user interface may include buttons, and optionally the user interface may further include a standard wired interface and a wireless interface.

[0065] Among them, the network interface may but is not limited to including a Bluetooth module, an NFC module, a Wi-Fi module, etc.

[0066] Among them, the processor may include one or more processing cores. The processor uses various interfaces and lines to connect various parts within the entire electronic device, and by running or executing instructions, programs, code sets, or instruction sets stored in the memory, and by calling data stored in the memory, it executes various functions of the routing device and processes data. Optionally, the processor may be implemented in at least one of the hardware forms of DSP, FPGA, and PLA. The processor may integrate one or several combinations of CPU, GPU, and modem, etc. Among them, the CPU mainly processes the operating system, user interface, application programs, etc.; the GPU is responsible for rendering and drawing the content that needs to be displayed on the display screen; the modem is used to process wireless communication. It can be understood that the above-mentioned modem may not be integrated into the processor and may be implemented separately by a single chip.

[0067] Among them, the memory may include RAM and may also include ROM. Optionally, the memory includes a non-transitory computer-readable medium. The memory can be used to store instructions, programs, codes, code sets, or instruction sets. The memory may include a program storage area and a data storage area. Among them, the program storage area may store instructions for implementing the operating system, instructions for at least one function (such as touch function, sound playback function, image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store the data involved in the above-mentioned various method embodiments. Optionally, the memory may further be at least one storage device located far from the aforementioned processor. As a computer storage medium, the memory may include an operating system, a network communication module, a user interface module, and application programs. The processor can be used to call the application programs stored in the memory and execute the methods in the above-mentioned one or more embodiments.

[0068] The embodiments of this specification also provide a computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are run on a computer or a processor, the computer or the processor is caused to execute one or more steps in the above method embodiments. If each component module of the above electronic device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in the computer-readable storage medium.

[0069] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the embodiments of this specification are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted through the computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center in a wired manner (such as coaxial cable, optical fiber, Digital Subscriber Line (DSL)) or a wireless manner (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that integrates one or more available media. The available medium can be a magnetic medium (such as a floppy disk, a hard disk, a magnetic tape), an optical medium (such as a Digital Versatile Disc (DVD)), or a semiconductor medium (such as a Solid State Disk (SSD)), etc.

[0070] Those of ordinary skill in the art can understand that all or part of the processes of implementing the methods in the above embodiments can be completed by instructing relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above various methods. The foregoing storage medium includes various media that can store program codes, such as ROM, RAM, magnetic disks, or optical discs. Without conflict, the technical features in this embodiment and the implementation solutions can be combined arbitrarily.

[0071] The embodiments described above are only described as the preferred embodiments of this specification, and do not limit the scope of this specification. Without departing from the design spirit of this specification, various deformations and improvements made by those of ordinary skill in the art to the technical solutions of this specification shall fall within the protection scope determined by the claims of this specification.

Claims

1. A differentiated resource allocation method based on user needs, comprising: Based on the user intention and the configuration strategy of the object, generating a first configuration combination, wherein the first configuration combination includes a plurality of objects of different types that meet the user intention and first configuration weights corresponding to the objects; Based on the adjacent discreteness, generating a plurality of second configuration combinations that are generalized and deviate from the first configuration combination, wherein the second configuration combination includes a plurality of objects of different types that meet the user's intention and second configuration weights corresponding to the objects; Based on the user input, acquiring a plurality of second configuration combinations related to the user input; as well as An optimal second configuration combination is selected therefrom to be output to the user.

2. The method according to claim 1, wherein generating the first configuration combination based on the user intention and the configuration strategy of the object comprises: Based on the object satisfying the user's intention obtained from the object pool, forming the first configuration combination consisting of a plurality of objects of different types satisfying the user's intention; as well as Based on the configuration strategy, the first configuration weight is configured for the object in the first configuration combination.

3. The method according to claim 2, wherein the object satisfying the user's intention obtained from the object pool comprises: Determine the user intention based on the user's historical profile, wherein the user intention includes object preference and user type; as well as Based on the user intent, an object satisfying the user intent is obtained from the object pool.

4. The method according to claim 2, wherein based on the configuration policy, configuring the first configuration weight for the object in the first configuration combination comprises: Obtaining sub-configuration strategies for different types of objects in the configuration strategy; With the goal of maximizing the total score of the first configuration weight in the first configuration combination, and the deviation between the first configuration weight of the first configuration combination and the benchmark configuration weight under different sub-configuration strategies is within a deviation range, generating the first configuration weight of the object in the first configuration combination; The first configuration combination is configured based on the generated first configuration weight.

5. The method according to claim 1 or 2, wherein generating the first configuration combination based on the user intention and the configuration strategy of the object comprises: Based on the objects at risk levels satisfying the user's intention obtained from the object pool, forming a first configuration combination consisting of a plurality of objects of different types; Based on a first strategy in the configuration strategy, determining a configuration ratio of the objects of the first type and the objects of other types in the first configuration combination at the risk level; as well as Based on the second strategy in the configuration strategy and the configuration proportion, a first configuration weight of the object of the first type and the objects of other types in the first configuration combination is determined.

6. The method according to claim 5, wherein determining the first configuration weights of the objects of the first type and the objects of other types in the first configuration combination based on the second strategy in the configuration strategy and the configuration proportion comprises: Obtaining sub-configuration policies for different types of the objects in the second policy; as well as Based on the sub-configuration strategy and the configuration proportion corresponding to the type of the object, a first configuration weight of the object of the type in the first configuration combination is determined, wherein the first configuration weight is within a range that fluctuates up and down based on the configuration proportion.

7. The method according to claim 6, wherein determining the first configuration weight of the object of the type in the first configuration combination based on the sub-configuration strategy corresponding to the type of the object and the configuration proportion comprises: With the goal of maximizing the total score of the first configuration weight in the first configuration combination, and the deviation between the first configuration weight of the first configuration combination and the benchmark configuration weight under different sub-configuration strategies is within a deviation range, the first configuration weight of the object of the type belonging to the first configuration combination is determined, wherein the first configuration weight is within a range of fluctuation based on the configuration proportion.

8. The method according to claim 6, wherein determining the first configuration weights of the objects of the first type and the objects of other types in the first configuration combination based on the second strategy in the configuration strategy and the configuration proportion further comprises: In response to the situation that the first configuration combination includes the object of the second type, based on the sub-configuration strategy of the object of the second type, the first configuration weight of the object of the second type is determined, and based on the sub-configuration strategy and the configuration ratio of the object of the first type and the objects of other types, the first configuration weights of the objects of the first type and the objects of other types are determined, wherein the first configuration weight is within a range that fluctuates up and down based on the configuration ratio.

9. The method according to claim 1, wherein generating a plurality of second configuration combinations deviating from the first configuration combination based on adjacent discreteness comprises: Taking the first configuration combination as a starting point, obtaining multiple second configuration combinations through multiple shifts; Each time the offset is performed, the second configuration combination under the current offset is determined with the goal of maximizing the adjacent discreteness between the second configuration combination under the current offset and the combination under the previous offset, and the current offset satisfies the first constraint condition of not exceeding the offset threshold.

10. The method according to claim 9, wherein for each shift, the goal is to maximize the adjacent discreteness between the second configuration combination under the current shift and the combination under the previous shift, and the current shift satisfies the first constraint condition of not exceeding the shift threshold, and determining the second configuration combination under the current shift comprises: Based on the first configuration combination, a first matrix corresponding to the first configuration combination is formed, and based on the second configuration combination, a second matrix corresponding to the second configuration combination is formed, and based on the offset, an offset matrix is ​​formed, wherein the first matrix includes objects of different types and the first configuration weights corresponding to the objects, the second matrix includes objects of different types and the second configuration weights corresponding to the objects, and the offset matrix includes objects of different types and the offset configuration weights corresponding to the objects; The offset matrix under the current offset is calculated by taking the maximization of adjacent discreteness between the second matrix under the current offset and the first matrix or the second matrix under the previous offset as the objective function, and the offset in the offset matrix under the current offset satisfies a first constraint condition that the offset does not exceed an offset threshold; and Based on the first matrix and the offset matrix at the current offset, the second matrix at the current offset is determined.

11. The method according to claim 9, wherein generating a plurality of second configuration combinations deviating from the first configuration combination based on adjacent discreteness comprises: Each time the shift is made, the second configuration combination under the current shift is determined with the goal of maximizing the adjacent discreteness between the second configuration combination under the current shift and the combination under the previous shift, and the current shift satisfies the first constraint condition of not exceeding the shift threshold, and the second configuration combination and the first configuration combination satisfy the second constraint condition of having the same industry exposure.

12. The method according to claim 11, wherein each time the offset is performed, the goal is to maximize the adjacent discreteness between the second configuration combination under the current offset and the combination under the previous offset, and the current offset satisfies the first constraint condition of not exceeding the offset threshold, and the second configuration combination and the first configuration combination satisfy the second constraint condition of the same industry exposure, and determining the second configuration combination under the current offset includes: Based on the first configuration combination, a first matrix corresponding to the first configuration combination is formed, and based on the second configuration combination, a second matrix corresponding to the second combination is formed, and based on the offset, an offset matrix is ​​formed, wherein the first matrix includes objects of different types and first configuration weights corresponding to the objects, the second matrix includes objects of different types and second configuration weights corresponding to the objects, and the offset matrix includes objects of different types and offset configuration weights corresponding to the objects; The offset matrix under the current offset is calculated by taking the maximization of the adjacent discreteness between the second matrix under the current offset and the first matrix or the second matrix under the previous offset as the objective function, and the offset amount in the offset matrix under the current offset satisfies a first constraint condition that does not exceed the offset threshold, and the second matrix under the current offset satisfies a second constraint condition that the industry exposure is the same as that of the first matrix; and Based on the first matrix and the offset matrix at the current offset, the second matrix at the current offset is determined.

13. A differentiated resource allocation device based on user needs, comprising: A first generating module is configured to generate a first configuration combination based on the user intention and the configuration strategy of the object, wherein the first configuration combination includes a plurality of objects of different types that meet the user intention and first configuration weights corresponding to the objects; A second generating module is configured to generate, based on the first configuration combination and adjacent discreteness, a plurality of second configuration combinations that are generalized deviations from the first configuration combination, wherein the second configuration combination includes a plurality of objects of different types that meet the user's intention and second configuration weights corresponding to the objects; and The output module is configured to obtain a plurality of second configuration combinations related to the user input based on the user input, and select an optimal second configuration combination therefrom to output to the user.

14. The apparatus according to claim 13, wherein the first generating module comprises: a forming unit configured to form a first configuration combination consisting of a plurality of objects of different types that meet the user's intention based on the objects that meet the user's intention obtained from the object pool; as well as A configuration unit is configured to configure the first configuration weight for the object in the first configuration combination based on the configuration strategy.

15. The device according to claim 14, wherein the first generating module further comprises: The intention extraction unit is configured to determine the user intention based on the user's historical portrait, and the user intention includes object preference and user type.

16. The device according to claim 13, wherein the second generating module comprises: An offset unit is configured to obtain a plurality of second configuration combinations by multiple offsets, starting from the first configuration combination; The generating unit is configured to determine the second configuration combination under the current offset each time with the goal of maximizing the adjacent discreteness between the second configuration combination under the current offset and the combination under the previous offset, and the current offset satisfies the first constraint condition of not exceeding the offset threshold.

17. Electronic equipment, including processor and memory; The processor is connected to the memory; The memory is used to store executable program code; The processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to execute the method according to any one of claims 1 to 12.

18. A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method according to any one of claims 1 to 12.

Images