Resource allocation method, apparatus, and device

Abstract

This specification discloses a resource allocation method, apparatus, and equipment. The scheme includes: identifying a resource allocation requester who holds a first resource and requires a second resource to supplement the first resource; identifying a resource holder who can provide the second resource and requires a third resource, wherein the third resource can be derived from the first and second resources; acquiring due diligence data related to the resource allocation requester and the resource holder; extracting a set of characteristic variables from the due diligence data, the set of characteristic variables including: a first subset of characteristic variables reflecting the resource holder's ability to provide the second resource; and a second subset of characteristic variables reflecting the resource allocation requester's ability to attract resources; generating risk measurement data for the first resource based on the first subset of characteristic variables; and generating derived expected data for the third resource based on the risk measurement data and the second subset of characteristic variables, so as to determine whether to allocate the second resource from the resource holder to the resource allocation requester based on the derived expected data.

Description

Technical Field

[0001] This specification relates to the field of Internet technology, and in particular to resource allocation methods, devices, and equipment. Background Technology

[0002] With the rapid development of computer technology and the internet, some offline businesses are gradually moving online. As the holder of computing resources to support online business operations, cloud computing needs to allocate its computing resources to different online business execution terminals during the online business execution process. Some terminals, after acquiring the computing resources allocated by the cloud, can execute online businesses more efficiently, increasing the volume of business executed and thus acquiring more traffic. Other terminals, however, execute the same volume of business as before the allocation, resulting in ineffective utilization of the computing resources allocated by the cloud and causing errors in the cloud computing resource allocation process.

[0003] Therefore, a more effective resource allocation plan is needed. Summary of the Invention

[0004] This specification provides one or more embodiments of a resource allocation method, apparatus, device, and storage medium to solve the following technical problem: the need for a more efficient resource allocation solution.

[0005] To solve the above-mentioned technical problems, one or more embodiments of this specification are implemented as follows:

[0006] This specification provides a resource allocation method according to one or more embodiments, including:

[0007] A resource allocation requester is identified who holds a first resource and requires a second resource to assist the first resource; a resource holder is identified who can provide the second resource and requires a third resource, wherein the third resource can be derived from the first resource and the second resource.

[0008] Obtain due diligence data related to the resource allocation requester and the resource holder;

[0009] Based on the due diligence data, a set of feature variables is extracted. The set of feature variables includes: a first subset of feature variables reflecting the resource holder's ability to provide the second resource, and a second subset of feature variables reflecting the resource allocation requester's ability to attract traffic.

[0010] Based on the first subset of feature variables, generate risk measurement data for the first resource;

[0011] Based on the risk measurement data and the second subset of feature variables, derivative expected data for the third resource is generated, so as to determine whether to allocate the second resource from the resource holder to the resource allocation requester based on the derivative expected data.

[0012] This specification provides a resource allocation method according to one or more embodiments, including:

[0013] Identify the investors who hold traffic assets and require the allocation of funds to support the investment in the traffic assets; identify the investors who can provide the allocation of funds and require profits, which can be derived from the traffic assets and the allocation of funds.

[0014] Obtain the investment requester and related due diligence data of the investor;

[0015] Based on the due diligence data, a set of characteristic variables is extracted. The set of characteristic variables includes: a first subset of characteristic variables reflecting the investor's ability to provide the allocated funds, and a second subset of characteristic variables reflecting the investment requester's ability to attract investment.

[0016] Based on the first subset of feature variables, generate risk control data for the traffic assets;

[0017] Based on the risk control data and the second subset of feature variables, profit expectation data is generated to determine whether to transfer the allocated funds from the investor to the investment requester.

[0018] This specification provides a resource allocation device according to one or more embodiments, comprising:

[0019] The module determines a resource allocation requester that holds a first resource and requires a second resource to assist the first resource, and determines a resource holder that can provide the second resource and requires a third resource, wherein the third resource can be derived from the first resource and the second resource.

[0020] The acquisition module acquires due diligence data related to the resource allocation requester and the resource holder.

[0021] The extraction module extracts a set of feature variables based on the due diligence data. The set of feature variables includes: a first subset of feature variables reflecting the resource holder's ability to provide the second resource, and a second subset of feature variables reflecting the resource allocation requester's ability to attract traffic.

[0022] The generation module generates risk measurement data for the first resource based on the first subset of feature variables.

[0023] The judgment module generates derived expected data for the third resource based on the risk measurement data and the second subset of feature variables, so as to determine whether to allocate the second resource from the resource holder to the resource allocation requester based on the derived expected data.

[0024] This specification provides a resource allocation device according to one or more embodiments, comprising:

[0025] The module identifies investors who hold traffic assets and require funding to support those assets, as well as investors who can provide the funding and require profits derived from the traffic assets and the funding.

[0026] The acquisition module acquires the investment requester and related due diligence data.

[0027] The extraction module extracts a set of feature variables based on the due diligence data. The set of feature variables includes: a first subset of feature variables reflecting the investor's ability to provide the allocated funds, and a second subset of feature variables reflecting the investment requester's ability to attract investment.

[0028] The generation module generates risk control data for the traffic assets based on the first subset of feature variables.

[0029] The judgment module generates profit expectation data based on the risk control data and the second subset of feature variables, so as to determine whether to transfer the allocated funds from the investor to the investment requester based on the profit expectation data.

[0030] This specification provides a resource allocation device according to one or more embodiments, comprising:

[0031] At least one processor; and,

[0032] A memory communicatively connected to the at least one processor; wherein,

[0033] The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enable the at least one processor to:

[0034] A resource allocation requester is identified who holds a first resource and requires a second resource to assist the first resource; a resource holder is identified who can provide the second resource and requires a third resource, wherein the third resource can be derived from the first resource and the second resource.

[0035] Obtain due diligence data related to the resource allocation requester and the resource holder;

[0036] Based on the due diligence data, a set of feature variables is extracted. The set of feature variables includes: a first subset of feature variables reflecting the resource holder's ability to provide the second resource, and a second subset of feature variables reflecting the resource allocation requester's ability to attract traffic.

[0037] Based on the first subset of feature variables, generate risk measurement data for the first resource;

[0038] Based on the risk measurement data and the second subset of feature variables, derivative expected data for the third resource is generated, so as to determine whether to allocate the second resource from the resource holder to the resource allocation requester based on the derivative expected data.

[0039] This specification provides a resource allocation device according to one or more embodiments, comprising:

[0040] At least one processor; and,

[0041] A memory communicatively connected to the at least one processor; wherein,

[0042] The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enable the at least one processor to:

[0043] Identify the investors who hold traffic assets and require the allocation of funds to support the investment in the traffic assets; identify the investors who can provide the allocation of funds and require profits, which can be derived from the traffic assets and the allocation of funds.

[0044] Obtain the investment requester and related due diligence data of the investor;

[0045] Based on the due diligence data, a set of characteristic variables is extracted. The set of characteristic variables includes: a first subset of characteristic variables reflecting the investor's ability to provide the allocated funds, and a second subset of characteristic variables reflecting the investment requester's ability to attract investment.

[0046] Based on the first subset of feature variables, generate risk control data for the traffic assets;

[0047] Based on the risk control data and the second subset of feature variables, profit expectation data is generated to determine whether to transfer the allocated funds from the investor to the investment requester.

[0048] This specification provides one or more embodiments of a non-volatile computer storage medium storing computer-executable instructions, wherein the computer-executable instructions are configured as follows:

[0049] A resource allocation requester is identified who holds a first resource and requires a second resource to assist the first resource; a resource holder is identified who can provide the second resource and requires a third resource, wherein the third resource can be derived from the first resource and the second resource.

[0050] Obtain due diligence data related to the resource allocation requester and the resource holder;

[0051] Based on the due diligence data, a set of feature variables is extracted. The set of feature variables includes: a first subset of feature variables reflecting the resource holder's ability to provide the second resource, and a second subset of feature variables reflecting the resource allocation requester's ability to attract traffic.

[0052] Based on the first subset of feature variables, generate risk measurement data for the first resource;

[0053] Based on the risk measurement data and the second subset of feature variables, derivative expected data for the third resource is generated, so as to determine whether to allocate the second resource from the resource holder to the resource allocation requester based on the derivative expected data.

[0054] The above-described at least one technical solution adopted in one or more embodiments of this specification can achieve the following beneficial effects: After determining the resource holder and the resource allocation requester, by extracting a set of characteristic variables from the acquired resource due diligence data, the data verification process brought about by traditional resource allocation based on historical data is avoided, the steps in the resource allocation process are reduced, and the efficiency of resource allocation is improved. Risk measurement data and derived expectation data are generated through the set of characteristic variables, which can extract the core elements of the resource allocation process and provide decision support for the resource allocation plan. Furthermore, the risk measurement data of the first resource generated allows for a more rigorous risk level classification of the first resource, enabling the resource holder to understand the risk status of the first resource in a timely manner, thus ensuring the security of the resource allocation process. Simultaneously, the derived expectation data of the third resource generated determines whether to proceed with the resource allocation process, providing an objective decision-making basis for executing the resource allocation process, making resource allocation decisions more scientific. Attached Figure Description

[0055] To more clearly illustrate the technical solutions in the embodiments or prior art of this specification, the drawings used in the description of the embodiments or prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0056] Figure 1 A flowchart illustrating a resource allocation method provided in one or more embodiments of this specification;

[0057] Figure 2 A schematic diagram of a resource allocation process in an application scenario, provided for one or more embodiments of this specification;

[0058] Figure 3 A schematic diagram of the structure of a resource allocation device provided in one or more embodiments of this specification;

[0059] Figure 4 This is a schematic diagram of the structure of a resource allocation device provided for one or more embodiments of this specification. Detailed Implementation

[0060] This specification provides resource allocation methods, apparatus, devices, and storage media through its embodiments.

[0061] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this application.

[0062] Figure 1 This diagram illustrates a resource allocation method provided in one or more embodiments of this specification. The method can be applied to various business domains, such as internet finance, e-commerce, instant messaging, gaming, and government services. The process can be executed by computing devices specific to the domain (e.g., intelligent customer service servers or intelligent mobile terminals for e-commerce). Certain input parameters or intermediate results within the process can be manually adjusted to improve accuracy.

[0063] Figure 1 The process may include the following steps:

[0064] S102: Determine a resource allocation requester who holds a first resource and requires a second resource to assist the first resource, and determine a resource holder who can provide the second resource and requires a third resource.

[0065] In the resource allocation process, there are resource holders and resource allocation requesters. The resource allocation requester is the party holding the primary resource, which can be an online service executed by computers or a derivative of that service, such as website traffic. Simultaneously, the resource allocation requester needs secondary resources. These secondary resources differ from the primary resources; they can assist the primary resource. For example, the secondary resource could be cloud computing power, which supports the execution of online services. Furthermore, the resource allocation requester may possess secondary resources, but the amount they possess may be insufficient to meet their own needs, thus requiring them to request allocation from the resource holder who holds the secondary resources.

[0066] Furthermore, the resource holder is the party holding the second resource, and if the resource holder wants to acquire a third resource, the third resource can be derived from the first and second resources. Specifically, after the resource holder allocates its second resource to the resource allocation requester, the resource allocation requester uses the acquired second resource to assist or support the first resource, so that the first resource generates the third resource during its operation / implementation. For example, if the first resource is online business and the second resource is cloud computing power, with the support of cloud computing power, the online business can be executed better, and the online business execution terminal can obtain more access. At the same time, the cloud computing power can also be known to more users, that is, obtain more access.

[0067] Meanwhile, the resource allocation process can be understood as the process by which the resource holder and the resource allocation requester exchange their respective resources, or it can be understood as the process by which the resource allocation requester requests a second resource from the resource holder, and the resource holder allocates its second resource to the resource allocation requester based on the request. In both of these processes, the resource holder will obtain a third resource, meaning that the resource holder will not provide its second resource to others for free.

[0068] S104: Obtain due diligence data related to the resource allocation requester and the resource holder.

[0069] After identifying the resource holder and the resource allocation requester, due diligence data related to both is obtained. This due diligence data can include three parts: first, data related to the resource holder, such as data on the resource holder's risk control regarding the allocation of the second resource, and management data related to the allocation of the second resource; second, data related to the resource allocation requester, such as data on the first resource held by the requester, and data on the requester's risk control regarding the allocation of the first resource; and third, data related to the industries associated with the first and second resources, such as risk data from previous resource allocations in related industries, and historical allocation data reflecting the resource allocation situation in related industries.

[0070] The process of obtaining the aforementioned due diligence data can be carried out on the servers corresponding to the resource holder and the resource allocation requester, or on typical industry websites related to the resource holder and the resource allocation requester. The obtained data may be disorganized; therefore, after obtaining the data, it is necessary to organize and classify it according to the three categories mentioned above to obtain the due diligence data.

[0071] S106: Based on the due diligence data, extract the set of feature variables.

[0072] To avoid the drawbacks of traditional resource allocation decisions based on historical data, the due diligence data is processed after acquisition to obtain a set of feature variables. This processing may involve directly filtering and extracting data from the due diligence dataset as feature variables, or it may involve integrating multiple data points to summarize a single feature variable that reflects the summarized data.

[0073] Since the feature variable set is extracted from due diligence data, if the due diligence data includes data related to the resource allocation requester, then the feature variable set will contain elements related to the resource allocation requester; similarly, if the due diligence data includes data related to the resource holder, then the feature variable set will contain elements related to the resource holder. These elements are essentially individual feature variables. Specifically, feature variables related to the resource holder are added to the first feature variable subset, enabling the first feature variable subset to reflect the resource holder's ability to provide the second resource. Simultaneously, feature variables related to the resource allocation requester are added to the second feature variable subset, enabling the second feature variable subset to reflect the resource allocation requester's traffic generation ability.

[0074] Traditional resource allocation processes rely on historical resource data, requiring verification to confirm its authenticity, which makes the process cumbersome. In this embodiment, to avoid these issues, the due diligence data is processed to extract a set of characteristic variables. This set considers only core factors influencing the resource allocation process, specifically those related to the resource holder and those related to the resource allocation requester. The resource holder, as the provider of the second resource, focuses on the quantity, duration, and conditions of the second resource. Therefore, the first subset of characteristic variables reflects the resource holder's ability to provide the second resource. The resource allocation requester, seeking to exchange their first resource for the second, focuses on the quality of the first resource. Higher quality first resources increase the likelihood of obtaining the second resource, while lower quality decreases this likelihood. Therefore, the second subset of characteristic variables reflects the quality of the first resource. In one possible example of this specification, the resource allocation requester may be a platform with traffic that wants to use its own traffic to obtain the required resources. In this case, the level of traffic and its growth become the factors of concern. At this time, the second feature variable subset reflects the traffic-driving ability of the resource holder.

[0075] S108: Generate risk measurement data for the first resource based on the first subset of feature variables.

[0076] Because the first subset of characteristic variables reflects the resource holder's ability to provide the second resource, it must include at least data related to the second resource. Generating risk measurement data for the first resource based on this subset is to assess its risk profile and provide a reference for resource allocation. Measuring the risk profile of the first resource, besides using data related to the second resource, will likely also require data related to the first resource itself. In other words, generating risk measurement data for the first resource will utilize not only data related to the second resource but also data related to the first resource held by the resource allocation requester. For example, if the resource allocation requester is a platform with traffic, measuring the risk profile of traffic (the first resource) will require not only data related to the second resource allocated through traffic but also data related to the traffic itself.

[0077] Furthermore, generating risk measurement data can be understood as calculating risk measurement data based on the feature variables in the first feature variable subset, or it can be understood as calculating risk measurement data not only based on the feature variables in the first feature variable subset, but also using data related to the first resource itself.

[0078] S110: Generate the derived expected data of the third resource based on the risk measurement data and the second subset of feature variables.

[0079] The third resource is derived from the first and second resources. That is, it is generated during the process of the second resource assisting the first resource in its operation / implementation. Therefore, whether a third resource can be derived, and the amount of such a resource, becomes a key decision-making basis for whether the resource allocation process should proceed. The expected data for the derivation of the third resource can be understood as data related to its generation, such as whether the third resource can be generated and the amount generated.

[0080] There are many factors that influence the expected data of third-party resources, but one of the more critical factors is the risk measurement data of the first resource. This is because if the first resource has a significant risk, the derived third resource will also have a correspondingly significant risk. Another factor is the traffic-generating ability of the resource allocation requester. For example, if the resource allocation requester requests resource allocation through a platform with a certain amount of traffic, then when judging the expected data of third-party resources, it is necessary to consider the platform's traffic-generating ability, that is, whether the platform can attract and maintain traffic.

[0081] From this perspective, the expected derivative data of the third resource is related to both the risk measurement data of the first resource and the traffic-driving ability of the resource allocation requester. It seems that the expected derivative data of the third resource is not so related to the resource holder. However, in fact, the risk measurement data of the first resource is at least related to the resource holder's second resource. This means that the data of the resource holder will also indirectly affect the derivative process of the third resource.

[0082] based on Figure 1 In addition to the method described herein, this specification also provides some specific implementation schemes and extension schemes of this method, which will be further explained below.

[0083] In one or more embodiments of this specification, to avoid data authenticity issues arising from traditional resource allocation decision-making processes using historical data, this specification, after obtaining due diligence data from the resource allocation requester and resource holder, needs to process the due diligence data to obtain feature variables, construct a feature variable set, and conduct resource allocation decision-making through the feature variable set, providing objective and authentic data indicators for decision-making. Specifically, due diligence data may be disorganized. To facilitate the extraction of representative feature variables, the due diligence data is first sorted into data related to the resource allocation requester and data related to the resource holder. After sorting, the data related to the resource holder is filtered to select directly usable and representative feature variables. Furthermore, the data related to the resource holder can be structured to integrate several types of data, using a single feature variable to represent the integrated data. The directly filtered and structured feature variables are stored in a feature variable set, thus obtaining the first subset of feature variables related to the resource holder. Similarly, by filtering the data related to the resource allocation requester, we can select the directly usable and representative feature variables. We can also structure the data related to the resource allocation requester, integrate several types of data, and use a feature variable to represent the integrated data. We can store the directly filtered and structured feature variables in a feature variable set to obtain a second feature variable subset related to the resource allocation requester.

[0084] Furthermore, the elements in the first subset of characteristic variables are related to the resource holder and reflect the resource holder's ability to provide the second resource. Therefore, the first subset of characteristic variables includes at least the second resource supply period, the average second resource supply quantity, and the second resource derivative rate. The second resource supply period represents the timeframe for which the resource holder will provide the second resource to the resource allocation requester. This period can be a maximum duration, representing the longest time the resource holder can provide the second resource to the resource allocation requester, or it can be a periodic value, representing the historical timeframe for which the second resource has been provided to the resource allocation requester, i.e., the time the resource allocation requester has used the second resource. The average second resource supply quantity represents the quantity of the second resource provided by the resource holder to the resource allocation requester. This quantity can be the average supply quantity determined based on historical resource allocation data. Using this average supply as the total supply, the second resource is provided to the resource allocation requester in one go. Alternatively, the average supply of the second resource can also be a phased average, meaning that when the resource holder provides the second resource to the resource allocation requester, it uses this average supply as the standard, providing one average supply of the second resource to the resource allocation requester each time, thus providing the total supply of the second resource to the resource allocation requester in batches. The second resource derivation rate represents the efficiency of the second resource in generating a third resource. For example, it can be calculated by the ratio between the third resource generated within a second resource supply period and the second resource supply provided within that supply period.

[0085] Furthermore, the elements in the second subset of characteristic variables are related to the resource allocation requester and reflect the requester's ability to attract traffic. Therefore, the second subset of characteristic variables includes at least the cumulative customer base and the periodically newly added customer base. In one example of this specification, when the first resource held by the resource holder is traffic resources, and a traffic request is used to allocate the second resource, the quality of the traffic becomes a key consideration in the resource allocation process. Indicators reflecting traffic quality include the amount of traffic and the potential for further increases. Therefore, when the first resource is traffic resources, cumulative traffic and newly added traffic become important factors reflecting the ability to attract traffic. In the second subset of characteristic variables, the cumulative customer base can be the number of customers held by the resource allocation requester at the current point in time, and the periodically newly added customer base can be the number of newly added customers within a reference period, which can be one day, one week, or one month.

[0086] In one or more embodiments of this specification, to ensure the effectiveness of the resource allocation process—that is, to ensure that the resource allocation process can generate as many third resources as possible—the resource allocation process needs to consider not only the resource holder's ability to provide second resources and the resource allocation requester's ability to attract traffic, but also the risks of the resource allocation process. If the risks are too high, the resource allocation process may fail to generate third resources, or even cause the allocated second resources to become invalid. Therefore, the set of feature variables will also include a subset of third feature variables reflecting the risks of resource allocation. Similar to the extraction process of the first and second subsets of characteristic variables, the extraction process of the third subset of characteristic variables is as follows: In the sorted due diligence data, data related to allocation risk is screened to select directly usable and representative characteristic variables. Furthermore, data related to allocation risk, such as risk data related to resource holders, risk data related to resource allocation requesters, and risk data existing in historical resource allocation processes, can be structured to integrate several types of data. A single characteristic variable is used to represent the integrated data. The directly screened and structured characteristic variables are stored in a single characteristic variable set, thus obtaining the third subset of characteristic variables reflecting resource allocation risk.

[0087] Furthermore, the third subset of characteristic variables, reflecting the risk situation of resource allocation, can be used in conjunction with the first subset of characteristic variables to generate risk measurement data for the first resource. In other words, the risk measurement data for the first resource is not only related to the supply factors of the second resource in the first subset of characteristic variables, but also uses industry-related data or data related to the first resource itself during the calculation process. Additionally, the elements in the third subset of characteristic variables must include at least: current resource risk level, current industry risk level, and external sampling risk level. The current resource risk level reflects the risk level of the first resource during the resource allocation process. That is, the risk level of the first resource is used when calculating the risk measurement data for the first resource, which is essential. The current industry risk level represents the risk level of the industry in which the first and / or second resources involved in the resource allocation process reside; this can be understood as the average risk level corresponding to the first resource. The external sampling risk level can be determined by sampling several historical resource allocation processes, rating the risk of the sampled data, and determining the risk level. The sampled data can be historical resource allocation processes or several similar resources corresponding to the first resource.

[0088] In one or more embodiments of this specification, to ensure the accuracy of the risk measurement data and to ensure that it can truly and effectively reflect the risks involved in the resource allocation process of the first resource, the risk measurement data in this specification may include two parts: one is the internal risk measurement data corresponding to the first resource, and the other is the external risk measurement data corresponding to the first resource, which can be represented by the external sampling risk level. In the process of generating the internal risk measurement data corresponding to the first resource, at least a subset of the first characteristic variables will be used, that is, at least factors related to the supply of the second resource will be used. Specifically, when the supply period of the second resource is the overall supply period, the average supply quantity of the second resource adopts the overall average, that is, this average supply quantity is used as the total supply quantity of the second resource and provided to the resource allocation requester at once for use during the aforementioned overall supply period. At this time, at least based on the supply period, supply quantity, and derivative rate of the second resource, the current resource risk level and the current industry risk level can be calculated. Then, the internal risk measurement data of the first resource can be determined based on the current resource risk level and the current industry risk level. Since the second resource has not yet been provided to the resource allocation requester, in this case, the supply period, average supply quantity, and derivative rate of the second resource can be determined based on historical data. These historical data can be data generated during the historical allocation of the second resource, or data generated during the historical allocation of resources similar to the second resource.

[0089] When the second resource supply period is a phased period, representing the time since resources were provided to the resource allocation requester, the average second resource supply represents the supply amount provided to the resource allocation requester in each phase. If the aforementioned situation indicates that the second resource has not yet been provided to the resource allocation requester, then the current situation involves providing the second resource in phases, with several phases already provided. In this case, the second resource supply period, the average second resource supply, and the second resource derivative rate can be determined based on the second resource already provided to the resource allocation requester. Furthermore, the process for determining the internal risk measurement data of the first resource is as follows: Extract supply time nodes from the second resource supply period, such as one month or two months. Then, by combining the average second resource supply with these time nodes, calculate the current second resource supply amount already provided to the resource allocation requester. Identify the overdue amount of the second resource within the second resource supply. Then, segment the overdue amount according to the previously determined time nodes to obtain the overdue amount of the second resource corresponding to each time node; for example, obtain second resources overdue for one month and second resources overdue for two months. Finally, by calculating the ratio between the overdue amount of the second resource and the supply of the second resource at each time point, the current resource risk level and the current industry risk level corresponding to the first resource are obtained. For example, if the second resource has been provided to the resource allocation requester for two months, the ratio between the overdue amount of the second resource for two months and the supply of the second resource is used as the current industry risk level; the ratio between the overdue amount of the second resource for one month and the supply of the second resource is used as the current resource risk level. It is understandable that in the above calculation of the current resource risk level, in addition to using elements from the first subset of characteristic variables, factors related to the second resource itself are also used.

[0090] Furthermore, after obtaining the current industry risk level and the current resource risk level corresponding to the first resource, the internal risk measurement data of the first resource can be calculated based on these two risk level values. For example, the calculation process is as follows: based on the current resource risk level and the current industry risk level, a risk range is determined, and then based on the probability that the second resource to be provided to the resource allocation requester falls into this risk range, the internal risk measurement data of the first resource is determined.

[0091] In one or more embodiments of this specification, to simplify the determination process of risk measurement data as much as possible while ensuring the validity of the risk measurement data, the external risk measurement data corresponding to the first resource can be directly represented by the external sampling risk level. The process of determining the external sampling risk level is as follows: Obtain several historical resources corresponding to the first resource. These historical resources can be the first resource itself or resources similar to the first resource. They are called historical resources because these resources have undergone or are undergoing resource allocation processes. Sample these historical resources. The specific sampling method is not limited. For example, random sampling or stratified sampling can be used. Assign different weights to the sampled historical resources and score them. The scoring criteria are determined based on the risk situation of each historical resource in the resource allocation process it participated in. Finally, multiply the scores and weights of each extracted historical resource and sum the products to obtain the external sampling risk level corresponding to the first resource, that is, the external risk measurement data corresponding to the first resource.

[0092] Furthermore, the internal and external risk measurement data are combined to obtain the risk measurement data corresponding to the first resource, which characterizes the risk situation of the first resource in the resource allocation process. Specifically, the internal and external risk measurement data are multiplied by their respective preset weights, and the products are summed to obtain the risk measurement data of the first resource.

[0093] In one or more embodiments of this specification, to determine the expected derivative data of the third resource and thus decide whether to execute the resource allocation process, it is necessary not only to consider the risk measurement data of the first resource, but also, when it is determined that the risk situation of the first resource is acceptable, to consider the derivative situation of the first and second resources on the third resource. If the derivative situation is not optimistic, it means that the resource allocation process in which the first and second resources are about to participate will not generate many third resources, and the significance of the resource allocation process is relatively small. In one example of this specification, the derivative capability of the third resource is used to characterize the derivative situation of the first and second resources on the third resource when participating in resource allocation. In the process of determining the risk measurement data for the first resource, the supply data of the second resource used may be historical supply data from previous resource allocation processes, or it may be existing supply data of the second resource in the current resource allocation process. Similarly, when determining the derivative capability of the third resource, the supply data of the second resource used may also be either of the above two types. However, regardless of the case, the process of determining the derivative capability of the third resource is as follows: Since the second resource will incur some losses when participating in the resource allocation process, the more times the second resource participates in the allocation, the higher its failure rate. Here, the failure rate refers to the portion of the second resource that can no longer participate in the resource allocation process. The annualized failure rate of the second resource is determined by multiplying the previously determined risk measurement data with the annualized number of allocations of the second resource. It should be noted that the annualized number of allocations here refers to the number of times the second resource participates in the resource allocation process on an annual basis. Of course, it can also be statistically analyzed on a weekly, monthly, or 10-year basis, as long as it is a regular periodic statistical analysis. Then, the third resource's derivative capacity is calculated using the supply of the second resource, the derivative rate of the second resource, and the annualized failure rate of the second resource. It is understood that in the actual calculation process, not only the above variables may be used, but other variables may also be used. Different calculation methods will result in different calculation content. Therefore, this specification does not limit the specific calculation method for the third resource's derivative capacity.

[0094] In one or more embodiments of this specification, after obtaining the risk measurement data of the first resource and the derivative capability of the third resource, the derivative expected data of the third resource can be calculated using these two data. The derivative expected data of the third resource represents the derivative situation of the first resource and the second resource when participating in resource allocation. Specifically, a training dataset is first constructed, which contains multiple input features. These input features are the factors in the resource allocation process determined above, including at least: the derivative capability of the third resource, the risk measurement data of the first resource, feature variables in the first feature variable subset, and feature variables in the second feature variable subset. The training process begins using the constructed training dataset. First, a function that can predict the derivative expected data is preset. Then, the value of each input feature is predicted based on this function. According to the prediction result, the Gini coefficient of each input feature is calculated. Then, the smallest Gini coefficient is found among these multiple Gini coefficients, and the input feature corresponding to this smallest Gini coefficient is found. Using this input feature as the splitting point, the training dataset is split into two sub-training datasets. The above operation is repeated for each sub-training dataset until the stopping splitting condition is met. In one example of this specification, the stopping condition for splitting can be that the minimum Gini coefficient corresponding to the input feature is less than a preset value, or that the number of input features in the split sub-training dataset is less than a preset value. After the training dataset stops splitting, a prediction function is obtained. The loss function of this prediction function is calculated, and if the loss value is large, the above training process is repeated to eventually obtain multiple prediction functions. Because each prediction function can exist in the form of a tree during the above training process, and each leaf of the tree corresponds to the value score of an input feature, the sum of the value scores of a given input feature across multiple trees is the final score for that input feature. Finally, based on the scores of all input features, the derived expected data for the third resource is determined.

[0095] Furthermore, based on the determined expected data for the third resource, a decision is made as to whether to proceed with the resource allocation process, i.e., whether to allocate the second resource from the resource holder to the resource allocation requester. Specifically, if the expected data for the third resource's derivation is favorable, it indicates a high probability of the third resource's derivation, and the resource allocation process can be executed. However, if the expected data for the third resource's derivation is unfavorable, it indicates a low probability of the third resource's derivation, and executing the resource allocation process is meaningless, i.e., the resource allocation process is rejected.

[0096] In one or more embodiments of this specification, the resource allocation requester can be a business execution terminal, and the resource holder can be a cloud computing terminal. In this case, the first resource is the executed business, the second resource is the computing power of the cloud computing, and the third resource is the access volume. The risk expectation data of the first resource is the execution volume of the business, and the derivative expectation data of the third resource is the expected data of the access volume earned. Furthermore, the supply period of the second resource is the provision period of the cloud computing power, the average supply of the second resource is the amount of computing power provided by the cloud computing, the derivative rate of the second resource is the increased access volume brought about by the computing power of the cloud computing, the current resource risk level is the increased business volume brought about by the computing power of the cloud computing, the current industry risk level is the average increased business volume brought about by the computing power of the cloud computing, the external sampling risk level is the increased business volume during the historical deployment of cloud computing power, the cumulative customer base is the cumulative business execution volume, and the periodically added customer base is the periodically increased business volume.

[0097] In one or more embodiments of this specification, based on the same idea, the above scheme can also be applied to the field of investment forecasting. The resource allocation requester can be the investment requester, and the resource holder can be the investor. In this case, the first resource is a flow asset, the second resource is the allocation fund, and the third resource is profit. The risk expectation data of the first resource is the risk control data of the flow asset, and the derivative expectation data of the third resource is the profit expectation data of the profit. Furthermore, the supply period of the second resource is the provision period of the allocation fund, the average supply of the second resource is the average provision of the allocation fund, the derivative rate of the second resource is the profit generated by the allocation fund, the current resource risk level is the current risk level of the flow asset, the current industry risk level is the average risk level of the flow asset, the external sampling risk level is the historical risk level of the flow asset, the cumulative customer base is the cumulative flow, and the periodically new customer base is the periodically new flow.

[0098] Based on the preceding explanation, more intuitively, combined with Figure 2 Let's take a more specific look at the solution in one application scenario. In this scenario, the resource holder is the investor, the resource allocation requester is the investment requester, the first resource is traffic assets, the second resource is allocation funds, and the third resource is profit.

[0099] Figure 2 This diagram illustrates a resource allocation process in one or more embodiments of this specification, within a given application scenario. Figure 2As shown, the resource allocation process mainly includes three processes: input, structuring, and output. The input data includes: historical asset performance, operating system, risk control system, credit product system, external benchmarking, and industry data. Among them, historical asset performance includes asset due diligence data, such as asset size and asset quality; the operating system includes asset source assessment, such as customer acquisition channels and user management; the risk control system includes risk control team structure, risk decision-making system, and risk management system; the credit product system includes the placement platform, marketing platform, and strategy platform; external benchmarking includes external cross-analysis, such as funding needs, credit risk, and external ratings; and industry data includes market environment and relevant company reports.

[0100] Furthermore, the structuring process mainly includes three parts: first, structured characteristic variables, including: t (average term of credit products), l (average historical loan amount), r (annualized interest rate), x (asset risk level), E (current industry risk level), m (current size of outstanding loans), c (cumulative loan customer base in the current month), n (new customer base in the current month), and p (credit approval rate in the current month); second, a structured sample set, assuming a total of M assets, connected to N funding parties, with a profit duration of D, and the i-th funding party is labeled after accessing the j-th asset for operation d. Third, structure the target variables. The profit rate is the target variable y.

[0101] Furthermore, the output process includes asset risk rating F(t,l,r,x,μ,E), assessment of profitability quality V(m,c,n,p,F), and prediction of profitability Prob(y), ultimately outputting an investment decision.

[0102] In one or more embodiments of this specification, the main process for predicting whether investor i will be profitable after reaching the profit duration d after accessing traffic asset j is as follows:

[0103] (1) Asset risk rating:

[0104] a. Based on the loan product's term t, average amount l, and annualized interest rate r, obtain the industry's average asset risk level as μ(t,l,r) and variance as δ(t,l,r). The current asset's risk level is x(t,l,r). The p-value corresponding to the statistical measure is used as the internal rating of the asset I(t,l,r); where μ(t,l,r) and x(t,l,r) are the industry-standard Vintage30+ / 90+ non-performing loan ratio indicators, which are obtained based on historical data statistics.

[0105] b. External sampling tests assess asset risk, corresponding to n scores {S1, S2, S3...S}. n}, adjust all scores to the range (0,1), with smaller scores indicating lower asset risk. Based on the ranking in 'a', the weight set corresponding to each score is {α1, α2, α3...α}. n}, where α i ∈[0,1), and If i = 1, 2, ..., n, then the corresponding external asset rating

[0106] c. Combining the above internal asset ratings I(t,l,r) with a weight of λ1 and the external asset ratings E with a weight of λ2, a comprehensive score is generated, i.e., F(t,l,r,x,μ,E)=λ1*I(t,l,r)+λ2*E.

[0107] (2) Assess the quality of earnings:

[0108] a. Assuming the total loan amount is m(t,l,r), the annualized bad debt rate is b, the annualized interest rate is r, and other costs are C(m) (including funding costs and data costs related to the loan size), then the net profit margin is... The annualized bad debt rate is obtained by combining the asset risk rating F(t,l,r,x,μ,E) and the annualized turnover β, i.e., b=F(t,l,r,x,μ,E)*β.

[0109] b. By combining two indicators—loan volume and net profit margin—and benchmarking against the industry average, an assessment of historical profitability quality is provided.

[0110] (3) Predict whether it will be profitable:

[0111] Assume there are M assets, connected to N investors, with a profit duration of D. The profit of the i-th investor after investing in the j-th asset and operating it for a duration of d is denoted as .

[0112] a. Sample: As described above, the sample set is...

[0113] b. Objective: If Then y = 1, otherwise y = 0;

[0114] c. Features: Divided into product type (product term, amount, interest rate, etc.), loan scale (average monthly loan amount, volatility, monthly new customer base size, volatility, etc.), asset risk rating before asset access (using 1), historical profitability quality assessed before access (using 2), etc.

[0115] Based on the same idea, one or more embodiments of this specification also provide apparatus and devices corresponding to the above methods, such as... Figure 3 , Figure 4 As shown.

[0116] Figure 3 This specification provides a schematic diagram of the structure of a resource allocation device according to one or more embodiments, the device comprising:

[0117] The determination module 302 determines a resource allocation requester who holds a first resource and requires a second resource to assist the first resource, and determines a resource holder who can provide the second resource and requires a third resource, wherein the third resource can be derived from the first resource and the second resource;

[0118] The acquisition module 304 acquires due diligence data related to the resource allocation requester and the resource holder.

[0119] The extraction module 306 extracts a set of feature variables based on the due diligence data. The set of feature variables includes: a first subset of feature variables reflecting the resource holder's ability to provide the second resource, and a second subset of feature variables reflecting the resource allocation requester's ability to attract traffic.

[0120] The generation module 308 generates risk measurement data for the first resource based on the first subset of feature variables.

[0121] The judgment module 310 generates derived expected data for the third resource based on the risk measurement data and the second subset of feature variables, so as to determine whether to allocate the second resource from the resource holder to the resource allocation requester based on the derived expected data.

[0122] Optionally, the extraction module 306 performs structured processing or filtering on the resource holder-related data in the due diligence data to obtain the first subset of feature variables, wherein the first subset of feature variables includes at least: the second resource supply period, the average second resource supply quantity, and the second resource derivative rate; and / or,

[0123] The extraction module 306 performs structured processing or filtering on the data related to the resource allocation requester in the due diligence data to obtain the second feature variable subset, which includes at least: cumulative customer groups and periodically newly added customer groups.

[0124] Optionally, the set of characteristic variables further includes: a third subset of characteristic variables reflecting the risk of resource allocation;

[0125] The extraction module 306 performs structured processing or filtering on resource risk-related data in the due diligence data to obtain the third feature variable subset, which is used to generate the risk measurement data. The third feature variable subset includes at least: current resource risk level, current industry risk level, and external sampling risk level.

[0126] Optionally, the generation module 308,

[0127] Multiple supply time points are extracted based on the second resource supply period, and the current second resource supply is determined through the multiple supply time points based on the average second resource supply.

[0128] Obtain the overdue amount of the second resource under the current second resource supply, and divide the overdue amount of the second resource according to the multiple supply time nodes to obtain the overdue amount of the second resource corresponding to each supply time node;

[0129] Calculate the ratio between the overdue amount of the second resource corresponding to each supply time node and the current supply amount of the second resource, and determine the current resource risk level and the current industry risk level corresponding to the first resource based on the ratio.

[0130] Optionally, the generation module 308,

[0131] The internal risk measurement data of the first resource is calculated based on the current resource risk level and the current industry risk level.

[0132] Based on the internal risk measurement data and the external sampling risk level, the risk measurement data of the first resource is calculated using preset weights.

[0133] The external sampling risk level is determined through the following process: obtaining several historical resources corresponding to the first resource, sampling from the several historical resources, and calculating the external sampling risk level corresponding to the first resource by assigning different weights to each sampling result.

[0134] Optionally, the device further includes a first determining module 312.

[0135] Based on the risk measurement data and the annualized number of allocations corresponding to the second resource, the annualized failure rate of the second resource is determined.

[0136] The third resource generation capacity is determined by the current supply of the second resource, the second resource generation rate, and the annualized failure rate of the second resource.

[0137] Based on the risk measurement data and the derivative capabilities of the third resource, the expected derivative data of the third resource is generated.

[0138] Optionally, the judgment module 310,

[0139] The training dataset is constructed using the third resource derivative capability, the risk measurement data, the feature variables in the first feature variable subset, and the feature variables in the second feature variable subset as input features.

[0140] The input features in the training dataset are predicted to take values, and the Gini coefficient corresponding to each input feature is calculated based on the prediction results.

[0141] Among the Gini coefficients corresponding to each input feature, the smallest Gini coefficient is determined, and the training dataset is split according to the input feature corresponding to the smallest Gini coefficient, so as to allocate the input features in the training dataset to two sub-training datasets;

[0142] Repeat the above operation until the number of input features in the sub-training dataset is less than a predetermined threshold, and a training result is obtained.

[0143] By adjusting the predicted value based on a single training result, and repeating the above operation, multiple training results can be obtained in the end.

[0144] Based on the scores corresponding to each input feature in the multiple training results, the derived expected data of the third resource is determined.

[0145] Optionally, the resource allocation requester is a business execution terminal, and the resource holder is a cloud computing terminal;

[0146] The first resource is the business being executed, the second resource is the computing power of cloud computing, and the third resource is the number of visits. The risk expectation data of the first resource is the execution volume of the business, and the derivative expectation data of the third resource is the expected data for earning visits.

[0147] Optionally, the resource allocation requester is an investment requester, and the resource holder is an investor;

[0148] The first resource is traffic assets, the second resource is allocated funds, and the third resource is profit. The risk expectation data of the first resource is the risk control data of the traffic assets, and the derivative expectation data of the third resource is the profit expectation data of the profit.

[0149] In one or more embodiments of this specification, the apparatus includes:

[0150] The determination module 302 determines the investment requester who holds the traffic assets and needs to allocate funds to assist the traffic assets, and determines the investor who can provide the allocation funds and needs the profit, the profit being derived from the traffic assets and the allocation funds;

[0151] Module 304 acquires the investment requester and related due diligence data.

[0152] The extraction module 306 extracts a set of feature variables based on the due diligence data. The set of feature variables includes: a first subset of feature variables reflecting the investor's ability to provide the allocated funds, and a second subset of feature variables reflecting the investment requester's ability to attract investment.

[0153] The generation module 308 generates risk control data for the traffic assets based on the first subset of feature variables.

[0154] The judgment module 310 generates profit expectation data based on the risk control data and the second feature variable subset, so as to determine whether to transfer the allocated funds from the investor to the investment requester based on the profit expectation data.

[0155] Figure 4 This specification provides a schematic diagram of the structure of a resource allocation device according to one or more embodiments, the device comprising:

[0156] At least one processor; and,

[0157] A memory communicatively connected to the at least one processor; wherein,

[0158] The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enable the at least one processor to:

[0159] A resource allocation requester is identified who holds a first resource and requires a second resource to assist the first resource; a resource holder is identified who can provide the second resource and requires a third resource, wherein the third resource can be derived from the first resource and the second resource.

[0160] Obtain due diligence data related to the resource allocation requester and the resource holder;

[0161] Based on the due diligence data, a set of feature variables is extracted. The set of feature variables includes: a first subset of feature variables reflecting the resource holder's ability to provide the second resource, and a second subset of feature variables reflecting the resource allocation requester's ability to attract traffic.

[0162] Based on the first subset of feature variables, generate risk measurement data for the first resource;

[0163] Based on the risk measurement data and the second subset of feature variables, derivative expected data for the third resource is generated, so as to determine whether to allocate the second resource from the resource holder to the resource allocation requester based on the derivative expected data.

[0164] In one or more embodiments of this specification, the device includes:

[0165] At least one processor; and,

[0166] A memory communicatively connected to the at least one processor; wherein,

[0167] The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enable the at least one processor to:

[0168] Identify the investors who hold traffic assets and require the allocation of funds to support the investment in the traffic assets; identify the investors who can provide the allocation of funds and require profits, which can be derived from the traffic assets and the allocation of funds.

[0169] Obtain the investment requester and related due diligence data of the investor;

[0170] Based on the due diligence data, a set of characteristic variables is extracted. The set of characteristic variables includes: a first subset of characteristic variables reflecting the investor's ability to provide the allocated funds, and a second subset of characteristic variables reflecting the investment requester's ability to attract investment.

[0171] Based on the first subset of feature variables, generate risk control data for the traffic assets;

[0172] Based on the risk control data and the second subset of feature variables, profit expectation data is generated to determine whether to transfer the allocated funds from the investor to the investment requester.

[0173] Based on the same idea, one or more embodiments of this specification also provide a non-volatile computer storage medium corresponding to the above method, storing computer-executable instructions, wherein the computer-executable instructions are configured as follows:

[0174] A resource allocation requester is identified who holds a first resource and requires a second resource to assist the first resource; a resource holder is identified who can provide the second resource and requires a third resource, wherein the third resource can be derived from the first resource and the second resource.

[0175] Obtain due diligence data related to the resource allocation requester and the resource holder;

[0176] Based on the due diligence data, a set of feature variables is extracted. The set of feature variables includes: a first subset of feature variables reflecting the resource holder's ability to provide the second resource, and a second subset of feature variables reflecting the resource allocation requester's ability to attract traffic.

[0177] Based on the first subset of feature variables, generate risk measurement data for the first resource;

[0178] Based on the risk measurement data and the second subset of feature variables, derivative expected data for the third resource is generated, so as to determine whether to allocate the second resource from the resource holder to the resource allocation requester based on the derivative expected data.

[0179] In one or more embodiments of this specification, the computer-executable instructions are configured as follows:

[0180] Identify the investors who hold traffic assets and require the allocation of funds to support the investment in the traffic assets; identify the investors who can provide the allocation of funds and require profits, which can be derived from the traffic assets and the allocation of funds.

[0181] Obtain the investment requester and related due diligence data of the investor;

[0182] Based on the due diligence data, a set of characteristic variables is extracted. The set of characteristic variables includes: a first subset of characteristic variables reflecting the investor's ability to provide the allocated funds, and a second subset of characteristic variables reflecting the investment requester's ability to attract investment.

[0183] Based on the first subset of feature variables, generate risk control data for the traffic assets;

[0184] Based on the risk control data and the second subset of feature variables, profit expectation data is generated to determine whether to transfer the allocated funds from the investor to the investment requester.

[0185] The above description is merely one or more embodiments of this specification and is not intended to limit this specification. Various modifications and variations can be made to the one or more embodiments of this specification by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of one or more embodiments of this specification should be included within the scope of the claims of this specification.

Claims

1. A resource allocation method, comprising: A resource allocation requester is identified who holds a first resource and requires a second resource to assist the first resource; a resource holder is identified who can provide the second resource and requires a third resource, wherein the third resource can be derived from the first resource and the second resource. Obtain due diligence data related to the resource allocation requester and the resource holder; Based on the due diligence data, a set of characteristic variables is extracted. This set includes: a first subset of characteristic variables reflecting the resource holder's ability to provide the second resource; and a second subset of characteristic variables reflecting the resource allocation requester's ability to attract traffic. The first subset of characteristic variables includes at least: the second resource supply period, the average second resource supply amount, and the second resource derivative rate. The second subset of characteristic variables includes at least: the cumulative customer base and the periodically newly added customer base. The second resource derivative rate represents the efficiency with which the second resource generates the third resource. The customer base refers to the customer base of the resource allocation requester. Based on the first subset of feature variables, generate risk measurement data for the first resource; Based on the risk measurement data and the second subset of feature variables, the derived expected data of the third resource is generated, so as to determine whether to allocate the second resource from the resource holder to the resource allocation requester based on the derived expected data; The resource allocation requester is a business execution terminal, and the resource holder is a cloud computing terminal; The first resource is the business being executed, the second resource is the computing power of cloud computing, and the third resource is the number of visits. The risk expectation data of the first resource is the execution volume of the business, and the derivative expectation data of the third resource is the expected data for earning visits.

2. The method as described in claim 1, wherein extracting the feature variable set based on the due diligence data specifically includes: In the due diligence data, the data related to the resource holder is structured or filtered to obtain the first subset of feature variables; And / or, In the due diligence data, the data related to the resource allocation requester is structured or filtered to obtain the second feature variable subset.

3. The method as described in claim 1, wherein the set of feature variables further includes: A third subset of characteristic variables reflecting the risk of resource allocation; The step of extracting a set of feature variables based on the due diligence data specifically includes: In the due diligence data, the data related to resource risks are structured or filtered to obtain the third feature variable subset, which is used to generate the risk measurement data. The third feature variable subset includes at least: the current resource risk level, the current industry risk level, and the external sampling risk level. The current resource risk level reflects the risk level of the first resource during the resource allocation process; The current industry risk level refers to the risk level of the industry in which the first and / or second resources involved in the resource allocation process are located. The external sampling risk level is determined by sampling several historical resource allocation processes, rating the risk of the sampled data, and establishing the risk level.

4. The method as described in claim 2, wherein generating risk measurement data for the first resource based on the first subset of feature variables specifically includes: Multiple supply time points are extracted based on the second resource supply period, and the current second resource supply is determined through the multiple supply time points based on the average second resource supply. Obtain the overdue amount of the second resource under the current second resource supply, and divide the overdue amount of the second resource according to the multiple supply time nodes to obtain the overdue amount of the second resource corresponding to each supply time node; Calculate the ratio between the overdue amount of the second resource corresponding to each supply time node and the current supply amount of the second resource, and determine the current resource risk level and the current industry risk level corresponding to the first resource based on the ratio.

5. The method of claim 3, wherein generating risk measurement data for the first resource based on the first subset of feature variables further comprises: The internal risk measurement data of the first resource is calculated based on the current resource risk level and the current industry risk level. Based on the internal risk measurement data and the external sampling risk level, the risk measurement data of the first resource is calculated using preset weights. The external sampling risk level is determined through the following process: obtaining several historical resources corresponding to the first resource, sampling from the several historical resources, and calculating the external sampling risk level corresponding to the first resource by assigning different weights to each sampling result.

6. The method of claim 4, before generating the derived expected data of the third resource based on the risk measurement data and the second subset of feature variables, the method further includes: Based on the risk measurement data and the annualized number of allocations corresponding to the second resource, the annualized failure rate of the second resource is determined. The third resource generation capacity is determined by the current supply of the second resource, the second resource generation rate, and the annualized failure rate of the second resource. Based on the risk measurement data and the derivative capabilities of the third resource, the expected derivative data of the third resource is generated.

7. The method of claim 6, wherein generating the expected derivative data of the third resource based on the risk measurement data and the derivative capability of the third resource specifically includes: The training dataset is constructed using the third resource derivative capability, the risk measurement data, the feature variables in the first feature variable subset, and the feature variables in the second feature variable subset as input features. The input features in the training dataset are predicted to take values, and the Gini coefficient corresponding to each input feature is calculated based on the prediction results. Among the Gini coefficients corresponding to each input feature, the smallest Gini coefficient is determined, and the training dataset is split according to the input feature corresponding to the smallest Gini coefficient, so as to allocate the input features in the training dataset to two sub-training datasets; The above operation is repeated for each of the sub-training datasets until the number of input features in the sub-training dataset is less than a predetermined threshold, thus obtaining a training result. By adjusting the predicted value based on a single training result, and repeating the above operation, multiple training results can be obtained in the end. Based on the scores corresponding to each input feature in the multiple training results, the derived expected data of the third resource is determined.

8. A resource allocation device, comprising: The module determines a resource allocation requester that holds a first resource and requires a second resource to assist the first resource, and determines a resource holder that can provide the second resource and requires a third resource, wherein the third resource can be derived from the first resource and the second resource. The acquisition module acquires due diligence data related to the resource allocation requester and the resource holder. The extraction module extracts a set of feature variables based on the due diligence data. This set includes: a first subset of feature variables reflecting the resource holder's ability to provide the second resource; and a second subset of feature variables reflecting the resource allocation requester's ability to attract traffic. The first subset of feature variables includes at least: the second resource supply period, the average second resource supply amount, and the second resource derivative rate. The second subset of feature variables includes at least: the cumulative customer base and the periodically newly added customer base. The second resource derivative rate represents the efficiency with which the second resource generates the third resource. The customer base refers to the customer base of the resource allocation requester. The generation module generates risk measurement data for the first resource based on the first subset of feature variables. The judgment module generates derived expected data for the third resource based on the risk measurement data and the second subset of feature variables, so as to determine whether to allocate the second resource from the resource holder to the resource allocation requester based on the derived expected data; The resource allocation requester is a business execution terminal, and the resource holder is a cloud computing terminal; The first resource is the business being executed, the second resource is the computing power of cloud computing, and the third resource is the number of visits. The risk expectation data of the first resource is the execution volume of the business, and the derivative expectation data of the third resource is the expected data for earning visits.

9. The apparatus as claimed in claim 8, The extraction module performs structured processing or filtering on the resource holder-related data in the due diligence data to obtain the first subset of feature variables; and / or, The extraction module performs structured processing or filtering on the data related to the resource allocation requester in the due diligence data to obtain the second feature variable subset.

10. The apparatus of claim 8, wherein the set of feature variables further comprises: A third subset of characteristic variables reflecting the risk of resource allocation; The extraction module performs structured processing or filtering on resource risk-related data in the due diligence data to obtain the third feature variable subset, which is used to generate the risk measurement data. The third feature variable subset includes at least: current resource risk level, current industry risk level, and external sampling risk level. The current resource risk level reflects the risk level of the first resource during the resource allocation process; The current industry risk level refers to the risk level of the industry in which the first and / or second resources involved in the resource allocation process are located. The external sampling risk level is determined by sampling several historical resource allocation processes, rating the risk of the sampled data, and establishing the risk level.

11. The apparatus of claim 9, wherein the generation module, Multiple supply time points are extracted based on the second resource supply period, and the current second resource supply is determined through the multiple supply time points based on the average second resource supply. Obtain the overdue amount of the second resource under the current second resource supply, and divide the overdue amount of the second resource according to the multiple supply time nodes to obtain the overdue amount of the second resource corresponding to each supply time node; Calculate the ratio between the overdue amount of the second resource corresponding to each supply time node and the current supply amount of the second resource, and determine the current resource risk level and the current industry risk level corresponding to the first resource based on the ratio.

12. The apparatus of claim 10, wherein the generation module, The internal risk measurement data of the first resource is calculated based on the current resource risk level and the current industry risk level. Based on the internal risk measurement data and the external sampling risk level, the risk measurement data of the first resource is calculated using preset weights. in, The external sampling risk level is determined through the following process: obtaining several historical resources corresponding to the first resource, sampling from the several historical resources, and calculating the external sampling risk level corresponding to the first resource by assigning different weights to each sampling result.

13. The apparatus of claim 11, further comprising a first determining module, Based on the risk measurement data and the annualized number of allocations corresponding to the second resource, the annualized failure rate of the second resource is determined. The third resource generation capacity is determined by the current supply of the second resource, the second resource generation rate, and the annualized failure rate of the second resource. Based on the risk measurement data and the derivative capabilities of the third resource, the expected derivative data of the third resource is generated.

14. The apparatus of claim 13, wherein the determining module, The training dataset is constructed using the third resource derivative capability, the risk measurement data, the feature variables in the first feature variable subset, and the feature variables in the second feature variable subset as input features. The input features in the training dataset are predicted to take values, and the Gini coefficient corresponding to each input feature is calculated based on the prediction results. Among the Gini coefficients corresponding to each input feature, the smallest Gini coefficient is determined, and the training dataset is split according to the input feature corresponding to the smallest Gini coefficient, so as to allocate the input features in the training dataset to two sub-training datasets; The above operation is repeated for each of the sub-training datasets until the number of input features in the sub-training dataset is less than a predetermined threshold, thus obtaining a training result. By adjusting the predicted value based on a single training result, and repeating the above operation, multiple training results can be obtained in the end. Based on the scores corresponding to each input feature in the multiple training results, the derived expected data of the third resource is determined.

15. A resource allocation device, comprising: At least one processor; as well as, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enable the at least one processor to: A resource allocation requester is identified who holds a first resource and requires a second resource to assist the first resource; a resource holder is identified who can provide the second resource and requires a third resource, wherein the third resource can be derived from the first resource and the second resource. Obtain due diligence data related to the resource allocation requester and the resource holder; Based on the due diligence data, a set of characteristic variables is extracted. This set includes: a first subset of characteristic variables reflecting the resource holder's ability to provide the second resource; and a second subset of characteristic variables reflecting the resource allocation requester's ability to attract traffic. The first subset of characteristic variables includes at least: the second resource supply period, the average second resource supply amount, and the second resource derivative rate. The second subset of characteristic variables includes at least: the cumulative customer base and the periodically newly added customer base. The second resource derivative rate represents the efficiency with which the second resource generates the third resource. The customer base refers to the customer base of the resource allocation requester. Based on the first subset of feature variables, generate risk measurement data for the first resource; Based on the risk measurement data and the second subset of feature variables, the derived expected data of the third resource is generated, so as to determine whether to allocate the second resource from the resource holder to the resource allocation requester based on the derived expected data; The resource allocation requester is a business execution terminal, and the resource holder is a cloud computing terminal; The first resource is the business being executed, the second resource is the computing power of cloud computing, and the third resource is the number of visits. The risk expectation data of the first resource is the execution volume of the business, and the derivative expectation data of the third resource is the expected data for earning visits.

Images