Virtual resource information processing method and device, electronic equipment and computer readable storage medium

Abstract

The invention discloses a virtual resource information processing method and device, electronic equipment and a computer readable storage medium, and the method comprises the steps: obtaining first historical transaction data of a first virtual resource, and enabling the first historical transaction data to comprise attribute information and real value information of the first virtual resource; the first transaction fluency of the first virtual resource is determined based on the first historical transaction data, the first transaction fluency is used for indicating the transaction fluency of the first virtual resource, and a value recommendation model is obtained through training based on the first historical transaction data and the first transaction fluency of the first virtual resource; determining predicted value information of the second virtual resource through a value recommendation model; acquiring second historical transaction data for performing a transaction on the second virtual resource based on the predicted value information; determining second transaction fluency of the second virtual resource based on the second historical transaction data; and updating the value recommendation model based on the second transaction fluency, so as to determine the value information of the virtual resources to be transacted based on the updated value recommendation model. The accuracy of the value information can be improved.

Description

Technical Field

[0001] This application relates to the field of game technologies, and in particular, to a method, apparatus, electronic device, and computer-readable storage medium for processing information of virtual resources. Background Art

[0002] In a game, a player can purchase items or sell the items they own. When selling the items they own, the price set for the sold items is usually recommended to the player by a game planner based on game experience.

[0003] However, the accuracy of the price determined based on game experience is low, resulting in the player having to list the items they own multiple times, with a long selling time, and reducing the liquidity of items in the game. Summary of the Invention

[0004] Embodiments of this application provide a method, apparatus, electronic device, and computer-readable storage medium for processing information of virtual resources, which can improve the accuracy of the price of items in a game, reduce the selling time of items in the game, and improve the liquidity of items in the game.

[0005] In a first aspect, embodiments of this application provide a method for processing information of virtual resources. The method includes:

[0006] Obtain first historical transaction data of a first virtual resource, where the first historical transaction data includes attribute information and true value information of the first virtual resource;

[0007] Determine a first transaction fluency of the first virtual resource based on the first historical transaction data, where the first transaction fluency is used to indicate the fluency of the transaction of the first virtual resource;

[0008] Train a value recommendation model based on the first historical transaction data and the first transaction fluency of the first virtual resource;

[0009] Determine predicted value information of a second virtual resource through the value recommendation model;

[0010] Obtain second historical transaction data of trading the second virtual resource based on the predicted value information;

[0011] Determine a second transaction fluency of the second virtual resource based on the second historical transaction data;

[0012] Update the value recommendation model based on the second transaction fluency, so as to determine value information of a virtual resource to be traded based on the updated value recommendation model.

[0013] In a second aspect, embodiments of this application further provide an apparatus for processing information of virtual resources. The apparatus includes:

[0014] A first acquisition module, configured to acquire first historical transaction data of a first virtual resource, where the first historical transaction data includes attribute information and true value information of the first virtual resource;

[0015] A first determination module, configured to determine a first transaction fluency of the first virtual resource based on the first historical transaction data, where the first transaction fluency is used to indicate the fluency degree of transactions of the first virtual resource;

[0016] A model training module, configured to train a value recommendation model based on the first historical transaction data and the first transaction fluency of the first virtual resource;

[0017] A second determination module, configured to determine predicted value information of a second virtual resource through the value recommendation model;

[0018] A second acquisition module, configured to acquire second historical transaction data of trading the second virtual resource based on the predicted value information;

[0019] A third determination module, configured to determine a second transaction fluency of the second virtual resource based on the second historical transaction data;

[0020] A model update module, configured to update the value recommendation model based on the second transaction fluency, so as to determine value information of a virtual resource to be traded based on the updated value recommendation model.

[0021] In a third aspect, an embodiment of the present application further provides an electronic device, including a memory storing multiple instructions; the processor loads the instructions from the memory to execute any information processing method of virtual resources provided by the embodiments of the present application.

[0022] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, where the computer-readable storage medium stores multiple instructions, and the instructions are suitable for being loaded by a processor to execute any information processing method of virtual resources provided by the embodiments of the present application.

[0023] In the embodiment of the present application, first historical transaction data of a first virtual resource is obtained. The first historical transaction data includes attribute information and real value information of the first virtual resource. Based on the first historical transaction data, a first transaction fluency of the first virtual resource is determined. The first transaction fluency is used to indicate the fluency degree of the transaction of the first virtual resource. Based on the first historical transaction data and the first transaction fluency of the first virtual resource, a value recommendation model is trained. The predicted value information of a second virtual resource is determined through the value recommendation model. Second historical transaction data of trading the second virtual resource based on the predicted value information is obtained. A second transaction fluency of the second virtual resource is determined based on the second historical transaction data. The value recommendation model is updated based on the second transaction fluency, so as to determine the value information of the virtual resource to be traded based on the updated value recommendation model. By training the model through the first historical transaction data and the first transaction fluency of the first virtual resource, the accuracy of the predicted value information determined through the value recommendation model is improved. Moreover, by updating the value model through the second transaction fluency, the accuracy of the value information of the virtual resource recommended by the updated value recommendation model is further improved, the selling time of virtual items in the game is reduced, and the liquidity of virtual items in the game is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative efforts.

[0025] Figure 1 It is a schematic diagram of a model processing system provided by an embodiment of the present application;

[0026] Figure 2 It is a schematic flowchart of an embodiment of a method for processing information of a virtual resource provided by an embodiment of the present application;

[0027] Figure 3 It is a schematic diagram of the value of summoned beast synthesis provided in an embodiment of the present application;

[0028] Figure 4 It is a schematic diagram of the value of embryos provided in an embodiment of the present application;

[0029] Figure 5 It is a schematic diagram of a preset table provided in an embodiment of the present application;

[0030] Figure 6 It is a schematic diagram of the weights corresponding to skills provided in an embodiment of the present application;

[0031] Figure 7It is a schematic diagram of the feature extraction process provided in the embodiments of the present application;

[0032] Figure 8 It is a schematic diagram of the value recommendation model generation provided in the embodiments of the present application;

[0033] Figure 9 It is a schematic diagram of the penalty rate and whether it is sold provided in the embodiments of the present application;

[0034] Figure 10 It is another schematic diagram of the penalty rate provided in the embodiments of the present application;

[0035] Figure 11 It is a schematic diagram of the second-sale rate provided in the embodiments of the present application;

[0036] Figure 12 It is a schematic diagram of the transaction smoothness provided in the embodiments of the present application;

[0037] Figure 13 It is a schematic diagram of the shelf interface provided in the embodiments of the present application;

[0038] Figure 14 It is another schematic diagram of the model processing system provided in the embodiments of the present application;

[0039] Figure 15 It is a schematic structural diagram of the information processing device for virtual resources provided in the embodiments of the present application;

[0040] Figure 16 It is a schematic structural diagram of the electronic device provided in the embodiments of the present application. Detailed implementation manners

[0041] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts shall fall within the protection scope of the present application.

[0042] Before explaining the embodiments of the present application in detail, some terms related to the embodiments of the present application will be explained.

[0043] Among them, in the description of the embodiments of the present application, terms such as "first" and "second" may be used herein to describe various concepts, but unless otherwise specified, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another. 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 does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices. "A plurality of" means two or more.

[0044] The embodiments of the present application provide a method, device, electronic device, and computer-readable storage medium for processing information of virtual resources. Specifically, the method for processing information of virtual resources in the embodiments of the present application can be executed by an electronic device, where the electronic device can be a terminal or a server and other devices.

[0045] The terminal can be a smart phone, a tablet computer, a laptop computer, a touch screen, a game console, a personal computer (PC), a personal digital assistant (PDA), and other terminal devices. The terminal can also include a client, and the client can be a game client, a browser client with a game program, or an instant messaging client, etc.

[0046] The server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers. It can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery network (CDN), and big data and artificial intelligence platforms.

[0047] For example, as Figure 1As shown in the figure, the electronic device is described by taking the terminal 10 and the server 20 as examples. The server 20 obtains the first historical transaction data of the first virtual resource. The first historical transaction data includes the attribute information and the real value information of the first virtual resource. Based on the first historical transaction data, the first transaction fluency of the first virtual resource is determined. The first transaction fluency is used to indicate the fluency degree of the transaction of the first virtual resource. Based on the first historical transaction data and the first transaction fluency of the first virtual resource, a value recommendation model is trained. The predicted value information of the second virtual resource is determined through the value recommendation model. The second historical transaction data of trading the second virtual resource based on the predicted value information is obtained. Based on the second historical transaction data, the second transaction fluency of the second virtual resource is determined. The value recommendation model is updated based on the second transaction fluency, so as to determine the value information of the virtual resource to be traded based on the updated value recommendation model. The server 20 receives the value recommendation request of the target virtual resource sent by the terminal 10. The server 20 recommends the value information of the target virtual resource through the updated value recommendation model based on the value recommendation request, and returns the value information of the target virtual resource to the terminal 10. The terminal 10 displays the value information of the target virtual resource.

[0048] The following will be described in detail with reference to the accompanying drawings respectively. It should be noted that the description order of the following embodiments does not limit the preferred order of the embodiments. Although the logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown in the drawings.

[0049] In this embodiment, the server is taken as an example for description. This embodiment provides a method for processing information of virtual resources, as Figure 2 shown, the specific process of the method for processing information of virtual resources may be as follows:

[0050] 201. Obtain the first historical transaction data of the first virtual resource. The first historical transaction data includes the attribute information and the real value information of the first virtual resource.

[0051] Among them, the virtual resource refers to the resources that the virtual character in the game can use in the virtual environment, which can decorate the virtual character, speed up its own attributes, assist in combat or initiate damage to other virtual characters. The type of virtual items can be set according to the actual situation. For example, the virtual resource may include at least one of virtual equipment, virtual props, and virtual accessories. The virtual equipment may include virtual weapons. The virtual props may include summoned beasts. The virtual accessories may include virtual clothing. This embodiment does not make any limitations here. The first virtual resource may be at least one of virtual items.

[0052] The first historical transaction data of the first virtual resource records a set of data on the behavior in the transaction of the first virtual resource. Optionally, the first historical transaction data may further include the transaction time, transaction data, and transaction status of the first virtual resource (the transaction status is used to indicate a successful transaction or a failed transaction). The attribute information of the first virtual resource is used to indicate the characteristics of the first virtual resource, and its type can be set according to the actual situation. For example, the attribute information of the first virtual resource may include at least one of the appearance attribute information, skill value attribute information, buff attribute information, and value attribute information of the first virtual resource. This embodiment does not make any limitations here.

[0053] When the first virtual resource is a summoned beast, the skill value attribute information is used to indicate at least one of the recommended skill combination value, weighted skill value, market skill value, and the value of the skill combination corresponding to different summoned beasts of the summoned beast. The buff attribute information is used to indicate the way to enhance the ability of the summoned beast. For example, the buff attribute information is used to indicate that enhancing the ability of the summoned beast requires "dragon bones", "lucky cakes", "ascension", "star ascension", "natural talent", "awakening", "growth rate", and "cultivation level". The value attribute information is used to indicate the value of the summoned beast. For example, the value attribute information is used to indicate the synthesis value of the summoned beast, the embryo value of the summoned beast, and the bottom line value of the summoned beast, etc.

[0054] Optionally, the synthesis value of the summoned beast can be determined based on the transaction value of the synthesis materials in the material market. The synthesis value of the summoned beast that is not directly synthesized is determined by the value of the peach blossom dew and the experience value. For example, the synthesis value of each summoned beast can be as Figure 3 shown.

[0055] Optionally, the embryo value of the summoned beast can be determined according to the historical transaction price of the recommended skills of the summoned beast. Specifically, the average value of the historical transaction prices that meet the preset conditions in the transaction data of the summoned beast with a transaction time of more than one year can be used as the embryo value of the summoned beast. For example, the preset condition is the latter 35%. The historical transaction prices in the transaction data of the summoned beast are C1, C2, C3, C4, C5, and C6, and the historical transaction prices of the latter 35% are C5 and C6. The average value of C5 and C6 is determined as the embryo value of the summoned beast. For example, the embryo value of the summoned beast can be as Figure 4 shown.

[0056] Optionally, the recommended skill combination value of the summoned beast can be determined according to the transaction prices of each skill of the summoned beast in the skill book. Specifically, when the skill of the summoned beast is in the preset table, it means that the skill is a recommended skill. The transaction prices of each skill of the summoned beast in the skill book can be added up to obtain the recommended skill combination value of the summoned beast. The preset table can be as Figure 5 shown. At this time, when the skill of the summoned beast is not in the preset table, the recommended skill combination value of the summoned beast is 0.

[0057] Optionally, the weighted skill value of the summoned beast can be obtained by multiplying the transaction price of the skills possessed by the summoned beast by the weight corresponding to the skill. For example, the skills possessed by the summoned beast and the weights corresponding to each skill can be as shown in Figure 6 , where Figure 6 skill_tieme in Figure 6 represents the skill book corresponding to the skill, Figure 6 skill_name in Figure 6 represents the name of the skill, Figure 6 skill_id in

[0058] represents the digital number corresponding to the skill,

[0059] rarity in

[0060] represents the type of the skill,

[0061] value_welig in

[0062] represents the weight corresponding to the skill.

[0063] The real value information of the first virtual resource is the value information determined artificially, which is used to indicate the virtual game assets required to purchase the first virtual resource. For example, if the virtual game assets are virtual game currency and the value information of the first virtual resource is 5 virtual game currency, it means that 5 virtual game currency is required to purchase the first virtual resource. Another example is that if the virtual game assets are game props and the value information of the first virtual resource is 2 game props of type A and 3 game props of type B, it means that 2 game props of type A and 3 game props of type B are required to purchase the first virtual resource.

[0059] Optionally, the first historical transaction data may include the transaction data of successful transactions or the transaction data of failed transactions.

[0060] 202. Determine the first transaction fluency of the first virtual resource based on the first historical transaction data, where the first transaction fluency is used to indicate the fluency of the transaction of the first virtual resource.

[0061] Among them, the transaction fluency (Transaction Fluency Score Index, TFSI) is used to indicate the fluency of the transaction of the virtual resource, so as to quantify the fluency of the transaction of the virtual resource. It can be determined according to at least one aspect of the selling time, selling rate, and whether the transaction is successful of the virtual resource. Since in the game, when the transaction is abnormal, the transaction may be withdrawn, indicating a failed transaction. Therefore, the transaction fluency can also be determined according to whether the transaction is successful. The first transaction fluency refers to the transaction fluency of the first virtual resource.

[0062] In some embodiments, determining the first transaction fluency of the first virtual resource based on the first historical transaction data includes:

[0063] Determine at least one of the first selling parameter and the first transaction risk coefficient of the first virtual resource based on the first historical transaction data;

[0064] Determine the first transaction fluency of the first virtual resource based on at least one of the first selling parameter and the first transaction risk coefficient.

[0065] Among them, the first selling parameter can be used to indicate the selling time and / or the selling quantity of the first virtual resource. When the first selling parameter is used to indicate the selling time of the first virtual resource, the first selling parameter can include a first selling time conversion coefficient (Time Transformation Coefficient, TTC). When the first selling parameter is used to indicate the selling quantity of the first virtual resource, the first selling parameter can include a first selling ratio. Optionally, the first selling ratio can be the first normalized ratio of sales (Normalized Ratio of Sales, NRS). Optionally, the first normalized ratio of sales can be calculated based on the selling time of the first virtual resource. The first transaction risk coefficient (Penalty Risk Ratio, PR) is used to indicate the risk of the transaction of the first virtual resource being penalized.

[0066] Optionally, the process of calculating the first normalized ratio of sales based on the selling time of the first virtual resource can be as follows: Based on the selling time of the first virtual resource, determine the selling quantity of the first virtual resource per time unit, and then based on the selling quantity of a certain time unit and the highest and lowest selling quantities within a preset duration, determine the first normalized ratio of sales. The preset duration includes multiple time units, and the time unit can be days or hours, which is not limited in this embodiment. Specifically, when the first normalized ratio of sales is used to indicate the normalized ratio of sales of a certain time unit, for example, if the time unit is days, the first normalized ratio of sales can be determined based on the selling quantity of a certain day and the highest and lowest selling quantities within a preset duration through formula (1):

[0067]

[0068] Among them, NRS represents the first normalized ratio of sales, n1 represents the selling quantity of a certain day, n2 represents the lowest selling quantity within the preset duration, n3 represents the highest selling quantity within the preset duration, and 0.001 represents a smoothing coefficient to avoid the first normalized ratio of sales being 0. The preset duration can be set according to the actual situation. For example, the preset duration can be 7 days or 8 days.

[0069] In this embodiment, based on the first historical transaction data, at least one of the first selling parameters and the first transaction risk coefficient of the first virtual resource is determined. Based on at least one of the first selling parameters and the first transaction risk coefficient, the first transaction fluency of the first virtual resource is determined, realizing the determination of the first transaction fluency through multiple dimensions. Compared with the method of directly calculating the first transaction fluency through the transaction success rate or the transaction failure rate, the accuracy of the first transaction fluency can be improved.

[0070] In some embodiments, the first selling parameter includes a first selling time conversion parameter. Based on the first historical transaction data, determining the first selling parameter of the first virtual resource includes:

[0071] Based on the first historical transaction data, determine the average selling time of the first virtual resource;

[0072] Based on the average selling time, determine the first selling time conversion parameter.

[0073] Among them, the average selling time of the first virtual resource can be determined based on the selling time in the first historical transaction data. The first logarithmic value can be determined based on the average selling time, and the first selling time conversion coefficient can be determined based on the first logarithmic value. Specifically, the average selling time can be substituted into formula (2) for calculation to obtain the first selling time conversion coefficient:

[0074]

[0075] Among them, TTC represents the first selling time conversion coefficient, log 10 (avgtime + 1) represents the first logarithmic value, and avgtime represents the average selling time.

[0076] In some embodiments, the first selling parameter includes a first selling time conversion parameter. Based on the first historical transaction data, determining the first selling parameter of the first virtual resource includes:

[0077] Based on the first historical transaction data, determine the average selling time of the first virtual resource;

[0078] Based on the average selling time and the time decay factor, determine the first selling time conversion parameter.

[0079] Among them, the first pair of numerical values can be determined based on the average selling time, and the first selling time conversion coefficient can be determined based on the first pair of numerical values and the time decay factor. Optionally, when the average selling time is greater than the first preset duration (the first preset duration can be set according to the actual situation, for example, the first preset duration can be 24 hours), the time decay factor can be determined based on the first pair of numerical values, and the first selling time conversion coefficient can be determined based on the first pair of numerical values and the time decay factor. When the average selling time is equal to or less than the first preset duration, the first selling time conversion coefficient can be determined based on the first pair of numerical values. Specifically, when the average selling time is equal to or less than the first preset duration, the first selling time conversion coefficient is determined by formula (2), and when the average selling time is greater than the first preset duration, the first selling time conversion coefficient is determined by formula (3):

[0080]

[0081] Among them, t represents the time decay factor, which can be calculated by formula (4):

[0082] t = e (-0.1×avgtime) (4)

[0083] When the average selling time is equal to or less than the first preset duration, the first selling time conversion coefficient is determined based on the first pair of numerical values. When the average selling time is greater than the first preset duration, the first selling time conversion coefficient is determined based on the first pair of numerical values and the time decay factor. This can make the proportion of the time conversion coefficient corresponding to the transaction data with a faster selling time in the transaction smoothness larger, improving the accuracy of the transaction smoothness.

[0084] In this embodiment, the first selling parameter includes the first selling time conversion parameter. Based on the first historical transaction data, the average selling time of the first virtual resource is determined, and based on the average selling time and the time decay factor, the first selling time conversion parameter is determined. This can make the proportion of the first selling time conversion parameter corresponding to the first historical transaction data with a faster selling time in the first transaction smoothness larger, improving the accuracy of the first smoothness.

[0085] In some embodiments, the first selling parameter includes the first transaction risk coefficient. Based on the first historical transaction data, determining the first transaction risk coefficient of the first virtual resource includes:

[0086] Based on the first historical transaction data, determine the number of failed transactions and the number of successful transactions;

[0087] Based on the number of failed transactions and the number of successful transactions, determine the first transaction risk coefficient of the first virtual resource.

[0088] Among them, in the game, when a transaction is abnormal, the transaction may be withdrawn, indicating that the transaction fails. Therefore, the number of failed transactions can be determined based on the first historical transaction data that has been withdrawn. Optionally, a first ratio can be determined based on the number of failed transactions and the number of successful transactions, and the smaller value between 1 and the first ratio is multiplied by the first risk amplification factor to obtain the first transaction risk coefficient. Specifically, the number of failed transactions and the number of successful transactions can be substituted into the above formula (5) for calculation to obtain the first transaction risk coefficient:

[0089]

[0090] Among them, PR represents the first transaction risk coefficient, f1 represents the number of failed transactions, f1 represents the number of successful transactions, and a1 represents the first risk amplification factor.

[0091] The first risk amplification factor can be determined by the number of failed transactions. Specifically, the first risk amplification factor can be calculated through formula (6):

[0092]

[0093] Among them, tanh() represents the hyperbolic tangent function.

[0094] Optionally, based on the first selling parameter and the first transaction risk coefficient, determine the first transaction smoothness of the first virtual resource, including:

[0095] Perform a weighted summation process on the first selling parameter and the first transaction risk coefficient to obtain the first transaction smoothness of the first virtual resource.

[0096] Among them, when the first selling parameter includes the first standard selling rate and the first selling time conversion coefficient of the first virtual resource, the first standard selling rate, the first selling time conversion coefficient, and the first transaction risk coefficient of the first virtual resource can be weighted and summed to obtain the first transaction smoothness. Specifically, the first standard selling rate, the first selling time conversion coefficient, and the first transaction risk coefficient of the first virtual resource can be substituted into formula (7) for calculation to obtain the first transaction smoothness:

[0097] TFSI = αNRS + βTTC + γ(1 - PR)(7)

[0098] Among them, TFSI represents the first transaction smoothness, α, β, and γ represent weight coefficients, and the sum of the three is 1.

[0099] In this embodiment, based on the first historical transaction data, the number of failed transactions and the number of successful transactions are determined. Based on the number of failed transactions and the number of successful transactions, the first transaction risk coefficient of the first virtual resource is determined, realizing the determination of the first transaction risk coefficient based on the number of failed transactions, so that the first transaction fluency can be determined through the number of failed transactions, making the first transaction fluency more accurate.

[0100] 203. Based on the first historical transaction data and the first transaction fluency of the first virtual resource, a value recommendation model is trained, and the first transaction fluency is used to indicate the fluency of the transaction of the first virtual resource.

[0101] Among them, the value recommendation model refers to a learning model with a value recommendation function, and its type can be set according to the actual situation. For example, the value recommendation model can be a decision tree or a neural network model. The neural network model can be, for example, a Convolutional Neural Networks (CNN) or a Long Short-Term Memory (LSTM). This embodiment does not make a limitation here.

[0102] Specifically, through the learning model to be trained, the attribute information in the first historical transaction data is subjected to feature extraction to obtain the extracted features. Based on the extracted features, the sample value information of the first virtual resource is predicted. Based on the sample value information and the first transaction fluency, the learning model to be trained is trained to obtain the value recommendation model.

[0103] In some embodiments, training the value recommendation model based on the first historical transaction data and the first transaction fluency of the first virtual resource includes:

[0104] Through the learning model to be trained, based on the first historical transaction data, the sample value information of the first virtual resource is predicted;

[0105] Based on the sample value information and the first historical transaction data, the sample transaction fluency of the first virtual resource is determined;

[0106] Based on the sample value information, the true value information, the sample transaction fluency and the first transaction fluency of the first virtual resource, the learning model is trained to obtain the value recommendation model.

[0107] Among them, through the learning model to be trained, the attribute information in the first historical transaction data can be feature-extracted to obtain the extracted features, and based on the extracted features, the sample value information of the first virtual resource can be predicted. The sample transaction fluency is used to quantitatively indicate the fluency of in-game players trading the first virtual resource based on the sample value information, and it can be determined according to at least one of the sample selling parameters and the sample transaction risk coefficient of the first virtual resource. Among them, the sample selling parameters can include at least one of the sample selling time and the sample selling rate, and the transaction risk coefficient can be determined by whether the transaction is successful.

[0108] Optionally, through the learning model to be trained, based on the sample value information and the first historical transaction data, the sample transaction fluency of the first virtual resource can be determined. Or, the sample transaction fluency of the first virtual resource can also be determined through the fluency model based on the sample value information and the first historical transaction data. This embodiment does not make a limitation here. Among them, the fluency model refers to a pre-trained learning model with the function of fluency prediction.

[0109] Optionally, determining the sample transaction fluency of the first virtual resource based on the sample value information and the first historical transaction data includes:

[0110] Determining at least one of the sample selling parameters and the sample transaction risk coefficient of the first virtual resource based on the sample value information and the first historical transaction data;

[0111] Determining the sample transaction fluency of the first virtual resource based on at least one of the sample selling parameters and the sample transaction risk coefficient.

[0112] Among them, through the learning model to be trained or the fluency model, at least one of the sample selling parameters and the sample transaction risk coefficient of the first virtual resource can be determined based on the sample value information and the first historical transaction data.

[0113] Optionally, the sample selling parameters can include at least one of the sample standard selling rate and the sample selling time conversion coefficient. The sample selling time of the first virtual resource can be determined based on the sample value information and the first historical transaction data. The sample standard selling rate can be calculated based on the sample selling time of the first virtual resource, and the average sample selling time of the sample selling time can be determined. The sample selling time conversion coefficient (Time Transformation Coefficient, TTC) can be determined based on the average sample selling time.

[0114] The number of successful sample transactions and the number of failed sample transactions of the first virtual resource can be determined based on the sample value information and the first historical transaction data (a failed transaction means the transaction is withdrawn). The sample transaction risk coefficient (Penalty Risk Ratio, PR) can be determined based on the number of failed sample transactions and the number of successful sample transactions. Finally, the sample transaction fluency of the first virtual resource can be determined based on at least one of the sample standard sales rate, the sample sales time conversion coefficient, and the sample transaction risk coefficient.

[0115] Optionally, the sample sales rate can be represented by the sample standard sales rate (Normalized Ratio of Sales, NRS). Specifically, the process of calculating the sample standard sales rate based on the sample sales time of the first virtual resource can be as follows: Through a learning model or a fluency model to be trained, based on the sample sales time of the first virtual resource, determine the sample sales quantity of the first virtual resource per time unit, and then based on the sample sales quantity of a certain time unit and the highest and lowest sample sales quantities within a near preset duration, determine the sample standard sales rate. The preset duration includes multiple time units, and the time unit can be days or hours, which is not limited in this embodiment. Specifically, when the sample standard sales rate is used to indicate the standard sales rate of a certain time unit, for example, if the time unit is days, the sample standard sales rate can be determined based on the sample sales quantity of a certain day and the highest and lowest sample sales quantities within the near preset duration through the above formula (1). At this time, NRS represents the sample standard sales rate, n1 represents the sample sales quantity of a certain day, n2 represents the lowest sample sales quantity within the preset duration, n3 represents the highest sample sales quantity within the preset duration, and 0.001 represents a smoothing coefficient to avoid the sample standard sales rate being 0. The preset duration can be set according to the actual situation. For example, the preset duration can be 7 days or 8 days.

[0116] Optionally, the process of determining the sample sales time conversion coefficient based on the average sample sales time can be as follows: Determine the sample logarithm value based on the average sample sales time, and determine the sample sales time conversion coefficient based on the sample logarithm value. Specifically, the average sample sales time can be substituted into the above formula (2) for calculation to obtain the sample sales time conversion coefficient. At this time, TTC represents the sample sales time conversion coefficient, log 10 (avgtime + 1) represents the sample logarithm value, and avgtime represents the average sample sales time.

[0117] Optionally, when the average sample selling time is greater than a first preset duration (the first preset duration can be set according to the actual situation. For example, the first preset duration can be 24 hours), the sample time decay factor can be determined based on the sample pair value, and the sample selling time conversion coefficient can be determined based on the sample pair value and the sample time decay factor. When the average sample selling time is equal to or less than the first preset duration, the sample selling time conversion coefficient is determined based on the sample pair value. Specifically, when the average sample selling time is equal to or less than the first preset duration, the sample selling time conversion coefficient is determined by formula (2). When the average sample selling time is greater than the first preset duration, the sample selling time conversion coefficient is determined by formula (3). At this time, t represents the sample time decay factor.

[0118] When the average sample selling time is equal to or less than the first preset duration, the sample selling time conversion coefficient is determined based on the sample pair value. When the average sample selling time is greater than the first preset duration, the sample selling time conversion coefficient is determined based on the sample pair value and the sample time decay factor. This can make the time conversion coefficient corresponding to the transaction data with a faster selling time have a greater proportion in the transaction fluency, improving the accuracy of the transaction fluency.

[0119] Optionally, the process of determining the sample transaction risk coefficient based on the sample transaction failure times and the sample transaction success times can be as follows: Determine the sample ratio based on the sample transaction failure times and the sample transaction success times, and multiply the smaller value of 1 and the sample ratio by the sample risk amplification coefficient to obtain the sample transaction risk coefficient. Specifically, the sample transaction failure times and the sample transaction success times can be substituted into the above formula (5) for calculation to obtain the sample transaction risk coefficient. At this time, PR represents the sample transaction risk coefficient, f1 represents the sample transaction failure times, f1 represents the sample transaction success times, and a1 represents the sample risk amplification coefficient.

[0120] The sample risk amplification coefficient can be determined by the sample transaction failure times. Specifically, the sample risk amplification coefficient can be calculated through the above formula (6).

[0121] After obtaining the standard selling rate of the first virtual resource, the sample selling time conversion coefficient, and the sample transaction risk coefficient, the standard selling rate of the first virtual resource, the sample selling time conversion coefficient, and the sample transaction risk coefficient can be weighted and summed to obtain the sample transaction fluency. Specifically, the standard selling rate of the first virtual resource, the sample selling time conversion coefficient, and the sample transaction risk coefficient can be substituted into the above formula (7) for calculation to obtain the sample transaction fluency. At this time, TFSI represents the sample transaction fluency, α, β, and γ represent the weight coefficients, and the sum of the three is 1.

[0122] Optionally, through the learning model to be trained, the sample value information of the first virtual resource can also be determined based on the player information of the first virtual resource and / or the server information where the player account is located. Among them, the player information is used to indicate the player's gaming ability, which can be set according to the actual situation. For example, the player information can include at least one of the game level of the player who owns the first virtual resource, the player's VIP information, and the player's combat power information. This embodiment does not make any limitations here. The server information can be set according to the actual situation. For example, the server information can include the opening duration of the server, the server skill value, and the server embryo value, etc. The server skill value can be the mean value of the skill value attribute information, and the server embryo value can be the mean value of the summoned beast embryo value.

[0123] When, through the learning model to be trained, feature extraction is performed on the attribute information of the first virtual resource, the player information of the first virtual resource, and the server information where the player account is located, and based on the extracted features, the sample value information of the first virtual resource is determined, the process of feature extraction can be as Figure 7 shown.

[0124] Optionally, after feature extraction is performed through the learning model to be trained, the extracted features can be further processed, and then based on the processed features, the sample value information of the first virtual resource is determined. At this time, after the learning model to be trained performs feature engineering processing on the attribute information, player information, and server information of the first virtual resource, the processed features are obtained. Feature engineering processing includes the processes of feature extraction and feature processing. The feature processing process can be set according to the actual situation. For example, the feature processing process can include at least one of the standardization processing of features (the standardization processing can be, for example, z-score processing or regular normalization processing), the missing value supplementation processing of features (the missing value supplementation processing can be, for example, continuous type supplementation processing and / or discrete value supplementation processing), the selection processing of features, the verification processing of features, the continuousization processing of features (the continuousization processing can be, for example, one-hot encoding processing), and the binning processing of features (the binning processing can be, for example, chi-square binning).

[0125] Among them, the process of the selection processing of features can be: through the learning model to be trained, determine the weight corresponding to each feature, multiply the weight corresponding to each feature by each feature, and obtain the selected features. The weight corresponding to each feature is used to represent the correlation between each feature and the sample value information and the importance of each feature in the process of determining the sample value information. The verification processing of features is used to verify the supplementation rate of features and / or whether the features are within the preset feature range. If the supplementation rate of features fails and / or the features are not within the preset feature range, feature extraction is performed again.

[0126] Optionally, after obtaining the value recommendation model, the value recommendation model can be evaluated using the data samples in the validation set. If the evaluation result of the value recommendation model meets the preset evaluation conditions, there is no need to adjust the value recommendation model anymore. At this time, the process of obtaining the value recommendation model can include preprocessing of the first historical transaction data, feature engineering processing, empirical rules, model training, model recommendation of sample value information, model parameter adjustment, and model evaluation. Among them, the empirical rules are used to set the maximum sample value information and the minimum value information. The feature engineering processing can include feature extraction and binning processing of features, as Figure 8 shown.

[0127] Optionally, the learning model to be trained can be trained using the XGBoost (eXtreme Gradient Boosting) algorithm. When training the learning model to be trained using the XGBoost algorithm, the number of leaf nodes of the decision tree, the maximum depth of the decision tree, the learning rate, the number of decision trees, the random seed, and the K-fold cross-validation strategy can be set. Setting the random seed is used to ensure the reproducibility of the results, and the K-fold cross-validation strategy is used to evaluate the performance and stability of the learning model.

[0128] In this embodiment, based on the first historical transaction data, the sample value information of the first virtual resource is predicted using the learning model to be trained. Based on the sample value information and the first historical transaction data, the sample transaction fluency of the first virtual resource is determined; based on the sample value information, the true value information, the sample transaction fluency, and the first transaction fluency of the first virtual resource, the learning model is trained to obtain the value recommendation model. When determining the sample transaction fluency of the first virtual resource based on the sample value information and the first historical transaction data using the learning model to be trained, multi-task learning is implemented to train the learning model, so as to assist the prediction of value information through the prediction of sample transaction fluency, and improve the accuracy of the value recommendation model when determining value information. When determining the sample transaction fluency of the first virtual resource based on the sample value information and the first historical transaction data using the fluency recognition model, the learning of the value information prediction by the learning model can be guided by the sample transaction fluency, and the accuracy of the value recommendation model when determining value information can be improved.

[0129] In some embodiments, training the learning model based on the sample value information, the true value information, the sample transaction fluency, and the first transaction fluency of the first historical transaction data to obtain the value recommendation model includes:

[0130] Determining a first loss function value based on the sample value information and the true value information;

[0131] Determining a second loss function value based on the sample transaction fluency and the first transaction fluency of the first virtual resource;

[0132] Train the learning model based on the first loss function value and the second loss function value to obtain a value recommendation model.

[0133] Among them, during the process of calculating the first loss function value, the true value information is used as a label, and during the process of calculating the second loss function, the first trading smoothness is used as a label. The first loss function and the second loss function can be weighted and summed to obtain the loss function value. If the loss function value indicates that the learning model converges, the learning model is determined as the value recommendation model. If the loss function value indicates that the learning model does not converge, update the model parameters of the learning model based on the loss function value, and return to execute the step of predicting the sample value information of the first virtual resource based on the first historical transaction data by the learning model to be trained.

[0134] Optionally, after obtaining the value recommendation model, the value recommendation model can be used to predict the verification value information based on the validation set, calculate the accuracy rate of the value recommendation model based on the verification value information, and evaluate the performance of the value recommendation model through the accuracy rate. The accuracy rate of the value recommendation model can be calculated by formula (8):

[0135]

[0136] Among them, r represents the accuracy rate of the value recommendation model, p1 represents the verification value information, and p2 represents the true value information corresponding to the validation set.

[0137] After experimental verification, the overall accuracy rate of the value information determined by the value recommendation model provided by this application can reach 90.2%. The selling rate of the virtual resources with the value information of the value recommendation model is 87.23%. The selling rate of the virtual resources without the value information of the value recommendation model is 11.59%, with a 75.64% increase. The penalty rate of the orders with the value information of the value recommendation model is reduced by 2.33% compared with the penalty rate of the orders without the value information of the value recommendation model. The specific data can be as Figure 9 and Figure 10 shown. The proportion of instant sell orders in the sold orders with the value information of the value recommendation model is reduced by 8.05% compared with the proportion of instant sell orders in the sold orders without the value information of the value recommendation model. The specific data can be as Figure 11 shown (the value information in the instant sell orders is inaccurate value information). After using the value recommendation model, the change in the trading smoothness can be as Figure 12 shown.

[0138] In this embodiment, based on the sample value information and the true value information, a first loss function value is determined. Based on the sample transaction fluency and the first transaction fluency of the first virtual resource, a second loss function value is determined. Based on the first loss function value and the second loss function value, a learning model is trained to obtain a value recommendation model, realizing the training of the learning model through multiple loss function values to further improve the accuracy of the value recommendation model.

[0139] 204. Determine the predicted value information of the second virtual resource through the value recommendation model.

[0140] Among them, the category of the second virtual resource and the category of the first virtual resource may be the same or different. Determining the predicted value information of the second virtual resource through the value recommendation model can avoid manually determining the predicted value information of the second virtual resource, improve the accuracy of the obtained value information of the second virtual resource, and thus enable the second virtual resource to be sold faster.

[0141] Optionally, the server can obtain a recommendation request for the second virtual resource. Through the value recommendation model, based on the recommendation request, determine the predicted value information of the second virtual resource, and return the predicted value information of the second virtual resource to the terminal. The terminal, in response to the confirmation operation of the predicted value information, lists the second virtual resource based on the predicted value information, so as to sell the second virtual resource through the predicted value information.

[0142] Optionally, the terminal can display a listing interface. The listing interface includes an editing area for the second virtual resource and the predicted value information. In response to the triggering operation on the editing area, obtain a recommendation request for the second virtual resource, and send the recommendation request to the server. The server, based on the recommendation request, determines the predicted value information of the second virtual resource through the value recommendation model, and returns the predicted value information to the terminal. The terminal displays a value information interface and displays the predicted value information of the second virtual resource determined through the value recommendation model on the value information interface. Among them, the value information interface can also have an input window, and the input window is used for players to manually input value information. The type of the triggering operation on the editing area can be set according to the actual situation. For example, the triggering operation on the editing area can be a click operation or a long-press operation. This embodiment does not make a limitation here.

[0143] For example, the listing interface can be as Figure 13 shown. In Figure 13 , the second virtual resource is a summoned beast of "Is love like a raging fire". After clicking on the editing area, a value information interface can be displayed. The editing area and the value information interface can be as Figure 13 shown. Figure 13 The reference price shown in is "472206784", and the reference price is the predicted value information of the second virtual resource determined through the value recommendation model.

[0144] Optionally, the first historical transaction data in the game can be stored in Kafka, and then the transaction data in Kafka will be consumed to store the first historical transaction data in Doris. The server obtains the first historical transaction data from Doris and trains a value recommendation model based on the first historical transaction data and the first transaction fluency. Then, it obtains the offline features of the second virtual resource from the Hive warehouse and inputs the offline features (the offline features can be, for example, the features extracted from the player information of the second virtual resource and / or the server information where the player account is located) and the second virtual resource into the value recommendation model. The predicted value information of the second virtual resource is obtained through the value recommendation model and output to the game client for display. For example, as Figure 14 shown, in Figure 14 , the tmax platform is the platform for deploying the value recommendation model.

[0145] 205. Obtain the second historical transaction data for trading the second virtual resource based on the predicted value information.

[0146] Among them, after the player obtains the predicted value information of the second virtual resource through the value recommendation model, the second virtual resource can be listed for sale so that the second virtual resource can be traded based on the predicted value information to obtain the second historical transaction data.

[0147] 206. Determine the second transaction fluency of the second virtual resource based on the second historical transaction data.

[0148] Among them, the method for calculating the second transaction fluency can refer to the method for calculating the first transaction fluency described above, and will not be elaborated in this embodiment.

[0149] 207. Update the value recommendation model based on the second transaction fluency to determine the value information of the virtual resource to be traded based on the updated value recommendation model.

[0150] Among them, since the second transaction fluency can be used to evaluate the recommendation performance of the value recommendation model, it is possible to determine whether to optimize the value recommendation model based on the second transaction fluency, so as to further improve the recommendation performance of the value recommendation model and further improve the accuracy of the recommended value information.

[0151] In some embodiments, updating the value recommendation model based on the second transaction fluency includes:

[0152] In the case where the second transaction fluency does not meet the preset fluency condition, update the value recommendation model based on the third historical transaction data. The third historical transaction data includes real transaction data, and the value information of the third virtual resource in the real transaction data is real value information.

[0153] Among them, the preset smoothness condition can be set according to the actual situation. For example, the preset smoothness condition can be equal to or greater than the preset smoothness degree, which is not limited in this embodiment. The third virtual resource and the second virtual resource can be the same or different. The true value information is the value information determined artificially.

[0154] The true transaction data in the third historical transaction data and the second historical transaction data can be data in the same time period. Optionally, the third historical transaction data can further include the second historical transaction data, so as to use the second historical transaction data as a negative sample to update the value recommendation model.

[0155] It can be understood that if the second historical transaction data includes transaction data for one time unit, when calculating the second transaction smoothness through the second selling rate for a certain time unit, a second transaction smoothness can be obtained. The second transaction smoothness is the smoothness of the second virtual resource in a certain time unit. If the second historical transaction data includes transaction data for at least two time units, when calculating the second transaction smoothness through the second selling rate for a certain time unit, multiple second transaction smoothnesses can be obtained. There is a corresponding second transaction smoothness for the transaction data of each time unit. At this time, when the preset number of second transaction smoothnesses do not meet the preset smoothness condition, the value recommendation model is updated based on the third historical transaction data.

[0156] In this embodiment, when the second transaction smoothness does not meet the preset smoothness condition, it indicates that the smoothness of the virtual resource transaction is low and the player's transaction experience is poor. Therefore, the value recommendation model can be updated based on the third historical transaction data. The third historical transaction data includes true transaction data, and the value information of the second virtual item in the true transaction data is the true value information, so as to improve the transaction smoothness, the transaction speed and the transaction success rate of the virtual resource, and reduce the penalty rate when trading based on the first value information recommended by the updated value recommendation model, thereby improving the player's transaction experience and further improving the player's retention rate on the platform.

[0157] In one implementation manner, updating the value recommendation model based on the third historical transaction data includes:

[0158] Through the value recommendation model, based on the attribute information of the third virtual resource in the third historical transaction data, determine the predicted value information of the third virtual resource;

[0159] Based on the predicted value information of the third virtual resource and the third historical transaction data, determine the third transaction smoothness of the third virtual resource;

[0160] Based on the predicted value information of the third virtual resource and the third transaction smoothness, determine the updated loss function value;

[0161] Update the value recommendation model based on the updated loss function value.

[0162] Among them, the attribute information of the third virtual resource is used to indicate the characteristics of the third virtual resource, and its type can be set according to the actual situation. For example, the attribute information of the third virtual resource can include at least one of the appearance attribute information, skill value attribute information, gain attribute information, and value attribute information of the third virtual resource. This embodiment does not make a limitation here. The definition of each attribute information of the third virtual resource can specifically refer to the definition of each attribute information of the first virtual resource above, and this embodiment will not elaborate here.

[0163] Optionally, through the value recommendation model, the predicted value information of the third virtual resource can also be determined based on the third virtual resource player information and / or the server information where the player account is located.

[0164] It can be understood that the process of determining the third transaction smoothness based on the predicted value information of the third virtual resource and the third historical transaction data can refer to the process of determining the sample transaction smoothness based on the sample value information and the first historical transaction data above, and this embodiment will not elaborate here.

[0165] Optionally, the process of determining the updated loss function value based on the predicted value information of the third virtual resource and the second transaction smoothness can be:

[0166] Determine the first updated loss function value based on the predicted value information of the third virtual resource;

[0167] Determine the second updated loss function value based on the third transaction smoothness;

[0168] Determine the updated loss function value based on the first updated loss function value and the second updated loss function value.

[0169] Among them, the first updated loss function value can be determined based on the predicted value information of the third virtual resource and the true value information in the third historical transaction data, the true third transaction smoothness can be calculated based on the third historical transaction data, the second updated loss function value can be determined based on the third transaction smoothness and the true third transaction smoothness, and the first updated loss function value and the second updated loss function value are weighted and summed to obtain the updated loss function value.

[0170] In this embodiment, through the value recommendation model, based on the attribute information of the third virtual resource in the third historical transaction data, the predicted value information of the third virtual resource is recommended. Based on the predicted value information and the third historical transaction data, the third transaction fluency of the third virtual resource is determined. Based on the predicted value information and the third transaction fluency, the updated loss function value is determined. Based on the updated loss function value, the value recommendation model is updated, so as to update the value recommendation model through multi-task learning, so as to assist the learning of the value information through the prediction of the third transaction fluency and improve the accuracy of the updated value recommendation model when determining the value information.

[0171] As can be seen from the above, in the embodiment of the present application, the first historical transaction data of the first virtual resource is obtained, and the first historical transaction data includes the attribute information and the true value information of the first virtual resource; based on the first historical transaction data, the first transaction fluency of the first virtual resource is determined, and the first transaction fluency is used to indicate the fluency of the transaction of the first virtual resource. Based on the first historical transaction data and the first transaction fluency of the first virtual resource, a value recommendation model is trained; through the value recommendation model, the predicted value information of the second virtual resource is determined; the second historical transaction data of trading the second virtual resource based on the predicted value information is obtained; based on the second historical transaction data, the second transaction fluency of the second virtual resource is determined; based on the second transaction fluency, the value recommendation model is updated, so as to determine the value information of the virtual resource to be traded based on the updated value recommendation model, realizing model training through the first historical transaction data and the first transaction fluency of the first virtual resource, improving the accuracy of the predicted value information determined by the value recommendation model, and further improving the accuracy of the value information of the virtual resource recommended by the updated value recommendation model by updating the value model through the second transaction fluency, reducing the selling time of virtual items in the game, and improving the liquidity of virtual items in the game.

[0172] To better implement the above method, the embodiment of the present application further provides an information processing device for virtual resources. The information processing device for virtual resources can be specifically integrated in an electronic device, such as a computer device, and the computer device can be a terminal, a server, or other devices.

[0173] Among them, the terminal can be a mobile phone, a tablet computer, a smart Bluetooth device, a laptop computer, a personal computer, or other devices; the server can be a single server or a server cluster composed of multiple servers.

[0174] For example, in this embodiment, taking the information processing device for virtual resources being specifically integrated in the terminal as an example, the method of the embodiment of the present application will be described in detail. This embodiment provides an information processing device for virtual resources, as Figure 15 shown, the information processing device for virtual resources may include:

[0175] The first acquisition module 1501 is configured to acquire first historical transaction data of a first virtual resource, where the first historical transaction data includes attribute information and true value information of the first virtual resource.

[0176] The first determination module 1502 is configured to determine a first transaction smoothness of the first virtual resource based on the first historical transaction data, where the first transaction smoothness is used to indicate the smoothness of the transaction of the first virtual resource.

[0177] The model training module 1503 is configured to train a value recommendation model based on the first historical transaction data and the first transaction smoothness of the first virtual resource, where the first transaction smoothness is used to indicate the smoothness of the transaction of the first virtual resource.

[0178] The second determination module 1504 is configured to determine predicted value information of a second virtual resource through the value recommendation model.

[0179] The second acquisition module 1505 is configured to acquire second historical transaction data of trading the second virtual resource based on the predicted value information.

[0180] The third determination module 1506 is configured to determine a second transaction smoothness of the second virtual resource based on the second historical transaction data.

[0181] The model update module 1507 is configured to update the value recommendation model based on the second transaction smoothness, so as to determine value information of a virtual resource to be traded based on the updated value recommendation model.

[0182] In some embodiments, the model update module 1507 is specifically configured to:

[0183] In the case that the second transaction smoothness does not meet a preset smoothness condition, update the value recommendation model based on third historical transaction data, where the third historical transaction data includes true historical transaction data, and the value information of a third virtual resource in the true historical transaction data is true value information.

[0184] In some embodiments, the model training module 1503 is specifically configured to:

[0185] Predict sample value information of the first virtual resource through a learning model to be trained, based on the first historical transaction data;

[0186] Determine a sample transaction smoothness of the first virtual resource based on the sample value information and the first historical transaction data;

[0187] Train the learning model based on the sample value information, the true value information, the sample transaction smoothness, and the first transaction smoothness of the first virtual resource to obtain the value recommendation model.

[0188] In some embodiments, the model training module 1503 is specifically configured to:

[0189] Based on the sample value information and the first historical transaction data, determine at least one of the sample selling parameters and the sample transaction risk coefficient of the first virtual resource;

[0190] Based on at least one of the sample selling parameters and the sample transaction risk coefficient, determine the sample transaction smoothness of the first virtual resource.

[0191] In some embodiments, the model training module 1503 is specifically configured to:

[0192] Based on the sample value information and the true value information, determine the first loss function value;

[0193] Based on the sample transaction smoothness and the first transaction smoothness of the first virtual resource, determine the second loss function value;

[0194] Based on the first loss function value and the second loss function value, train the learning model to obtain a value recommendation model.

[0195] In some embodiments, the first determination module 1502 is specifically configured to:

[0196] Based on the first historical transaction data, determine at least one of the first selling parameters and the first transaction risk coefficient of the first virtual resource;

[0197] Based on at least one of the first selling parameters and the first transaction risk coefficient, determine the first transaction smoothness of the first virtual resource.

[0198] In some embodiments, the first determination module 1502 is specifically configured to:

[0199] Perform a weighted summation process on the first selling parameter and the first transaction risk coefficient to obtain the first transaction smoothness of the first virtual resource.

[0200] In some embodiments, the first selling parameter includes a first selling time conversion parameter, and the first determination module 1502 is specifically configured to:

[0201] Based on the first historical transaction data, determine the average selling time of the first virtual resource;

[0202] Based on the average selling time and the time decay factor, determine the first selling time conversion parameter.

[0203] In some embodiments, the first selling parameter includes a first transaction risk coefficient, and the first determination module 1502 is specifically configured to:

[0204] Based on the first historical transaction data, determine the number of failed transactions and the number of successful transactions;

[0205] Determine a first transaction risk coefficient of the first virtual resource based on the number of failed transactions and the number of successful transactions.

[0206] In specific implementation, each of the above modules can be implemented as an independent entity, or can be arbitrarily combined and implemented as the same or several entities. For the specific implementation manners of each of the above modules and the corresponding beneficial effects, reference can be made to the foregoing method embodiments, which will not be elaborated herein.

[0207] Correspondingly, an embodiment of the present application further provides an electronic device, which may be a terminal, and the terminal may be a terminal device such as a smart phone, a tablet computer, a notebook computer, a touch screen, a game console, a personal computer (PC, Personal Computer), a personal digital assistant (Personal Digital Assistant, PDA), etc. As Figure 16 shown, Figure 16 is a schematic structural diagram of the electronic device provided by the embodiment of the present application. The electronic device 1600 includes a processor 1601 having one or more processing cores, a memory 1602 having one or more computer-readable storage media, and a computer program stored in the memory 1602 and executable on the processor. Among them, the processor 1601 is electrically connected to the memory 1602. Those skilled in the art can understand that the structural diagram of the electronic device shown in the figure does not constitute a limitation on the electronic device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange different components.

[0208] The processor 1601 is the control center of the electronic device 1600, connects various parts of the entire electronic device 1600 through various interfaces and lines, and executes various functions of the electronic device 1600 and processes data by running or loading software programs and / or modules stored in the memory 1602, and calling data stored in the memory 1602, so as to perform overall monitoring on the electronic device 1600.

[0209] In the embodiment of the present application, the processor 1601 in the electronic device 1600 will load instructions corresponding to the processes of one or more application programs into the memory 1602 according to the following steps, and the processor 1601 will run the application programs stored in the memory 1602 to implement various functions, such as:

[0210] Obtain first historical transaction data of the first virtual resource, where the first historical transaction data includes attribute information and true value information of the first virtual resource;

[0211] Determine the first transaction fluency of the first virtual resource based on the first historical transaction data, where the first transaction fluency is used to indicate the fluency of the transaction of the first virtual resource;

[0212] Train a value recommendation model based on the first historical transaction data and the first transaction fluency of the first virtual resource;

[0213] Determine the predicted value information of the second virtual resource through the value recommendation model;

[0214] Obtain the second historical transaction data of trading the second virtual resource based on the predicted value information;

[0215] Determine the second transaction fluency of the second virtual resource based on the second historical transaction data;

[0216] Update the value recommendation model based on the second transaction fluency, so as to determine the value information of the virtual resource to be traded based on the updated value recommendation model.

[0217] For the specific implementation manners of the above operations and the corresponding beneficial effects, reference may be made to the detailed description of the information processing method for virtual resources above, which will not be elaborated here.

[0218] Optionally, as Figure 16 shown, the electronic device 1600 further includes: a touch display screen 1603, a radio frequency circuit 1604, an audio circuit 1605, an input unit 1606, and a power supply 1607. Among them, the processor 1601 is electrically connected to the touch display screen 1603, the radio frequency circuit 1604, the audio circuit 1605, the input unit 1606, and the power supply 1607 respectively. Those skilled in the art can understand that Figure 16 the structure of the electronic device shown in

[0219] The touch display screen 1603 can be used to display a graphical user interface and receive operation instructions generated by a user acting on the graphical user interface. The touch display screen 1603 may include a display panel and a touch panel. Among them, the display panel can be used to display information input by the user or information provided to the user, as well as various graphical user interfaces of the electronic device. These graphical user interfaces can be composed of graphics, text, icons, videos, and any combination thereof. Optionally, the display panel can be configured in the form of a liquid crystal display (LCD), an organic light-emitting diode (OLED), etc. The touch panel can be used to collect touch operations of the user on or near it (such as operations of the user using a finger, a stylus, or any suitable object or accessory on or near the touch panel), and generate corresponding operation instructions, and the operation instructions execute the corresponding program. Optionally, the touch panel can include two parts: a touch detection device and a touch controller. Among them, the touch detection device detects the touch position of the user and detects the signal brought by the touch operation, and transmits the signal to the touch controller; the touch controller receives the touch information from the touch detection device, converts it into contact coordinates, and then sends it to the processor 1601, and can receive and execute the commands sent by the processor 1601. The touch panel can cover the display panel. After the touch panel detects a touch operation on or near it, it is transmitted to the processor 1601 to determine the type of touch event. Subsequently, the processor 1601 provides a corresponding visual output on the display panel according to the type of touch event. In the embodiment of the present application, the touch panel and the display panel can be integrated into the touch display screen 1603 to implement input and output functions. However, in some embodiments, the touch panel and the touch panel can be implemented as two independent components to implement input and output functions. That is, the touch display screen 1603 can also be used as a part of the input unit 1606 to implement the input function.

[0220] The radio frequency circuit 1604 can be used to transmit and receive radio frequency signals to establish wireless communication with a network device or other electronic devices through wireless communication, and transmit and receive signals with the network device or other electronic devices.

[0221] The audio circuit 1605 can be used to provide an audio interface between the user and the electronic device through a speaker and a microphone. The audio circuit 1605 can transmit the electrical signal converted from the received audio data to the speaker, which converts it into a sound signal for output; on the other hand, the microphone converts the collected sound signal into an electrical signal, which is received by the audio circuit 1605 and then converted into audio data. After the audio data is output and processed by the processor 1601, it is sent through the radio frequency circuit 1604 to, for example, another electronic device, or the audio data is output to the memory 1602 for further processing. The audio circuit 1605 may also include an earphone jack to provide communication between the peripheral earphone and the electronic device.

[0222] The input unit 1606 can be used to receive input digital, character information or user characteristic information (such as fingerprint, iris, face information, etc.), and generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function controls.

[0223] The power supply 1607 is used to supply power to each component of the electronic device 1600. Optionally, the power supply 1607 can be logically connected to the processor 1601 through a power management system, so as to implement functions such as management of charging, discharging, and power consumption management through the power management system. The power supply 1607 may also include any components such as one or more DC or AC power supplies, a recharge system, a power failure detection circuit, a power converter or inverter, and a power status indicator.

[0224] Although Figure 16 not shown in the figure, the electronic device 1600 may also include a camera, a sensor, a Wi-Fi module, a Bluetooth module, etc., which will not be elaborated here.

[0225] In the above embodiments, the descriptions of the respective embodiments have their own emphases. For parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0226] Those of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructions or by controlling relevant hardware through instructions. The instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0227] Therefore, the embodiments of the present application provide a computer-readable storage medium, in which multiple computer programs are stored. The computer programs can be loaded by a processor to execute any information processing method of virtual resources provided by the embodiments of the present application. For example, the computer program can execute the following steps:

[0228] Obtain the first historical transaction data of the first virtual resource, where the first historical transaction data includes the attribute information and the true value information of the first virtual resource;

[0229] Determine the first trading fluency of the first virtual resource based on the first historical transaction data, where the first trading fluency is used to indicate the fluency of the transaction of the first virtual resource;

[0230] Train a value recommendation model based on the first historical transaction data and the first trading fluency of the first virtual resource;

[0231] Determine the predicted value information of the second virtual resource through the value recommendation model;

[0232] Obtain the second historical transaction data for trading the second virtual resource based on the predicted value information;

[0233] Determine the second trading fluency of the second virtual resource based on the second historical transaction data;

[0234] Update the value recommendation model based on the second trading fluency, so as to determine the value information of the virtual resource to be traded based on the updated value recommendation model.

[0235] For the specific implementation manners of the above operations and the corresponding beneficial effects, reference may be made to the detailed description of the information processing method for virtual resources above, and details are not described herein again.

[0236] Wherein, the computer-readable storage medium may include: read-only memory (ROM, Read Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disc, etc.

[0237] Since the computer program stored in the computer-readable storage medium can execute any information processing method for virtual resources provided in the embodiments of the present application, therefore, the beneficial effects that can be achieved by any information processing method for virtual resources provided in the embodiments of the present application can be realized. For details, refer to the previous embodiments and will not be elaborated herein again.

[0238] The above has introduced in detail an information processing method, device, electronic device and computer-readable storage medium for virtual resources provided in the embodiments of the present application. Specific examples are used in this article to elaborate on the principle and implementation manner of the present application. The description of the above embodiments is only used to help understand the method and its core idea of the present application; at the same time, for those skilled in the art, according to the idea of the present application, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to the present application.

Claims

1. A method for processing information of virtual resources, characterized in that, The method includes: Obtaining first historical transaction data of a first virtual resource, where the first historical transaction data includes attribute information and true value information of the first virtual resource; Determining a first transaction fluency of the first virtual resource based on the first historical transaction data, where the first transaction fluency is used to indicate the fluency degree of transactions of the first virtual resource; Training a value recommendation model based on the first historical transaction data and the first transaction fluency of the first virtual resource; Determining predicted value information of a second virtual resource through the value recommendation model; Obtaining second historical transaction data of trading the second virtual resource based on the predicted value information; Determining a second transaction fluency of the second virtual resource based on the second historical transaction data; Updating the value recommendation model based on the second transaction fluency, so as to determine value information of a virtual resource to be traded based on the updated value recommendation model.

2. The method according to claim 1, characterized in that The updating the value recommendation model based on the second transaction fluency includes: In the case that the second transaction fluency does not meet a preset fluency condition, updating the value recommendation model based on third historical transaction data, where the third historical transaction data includes true historical transaction data, and the value information of a third virtual resource in the true historical transaction data is true value information.

3. The method according to claim 1, characterized in that, The training the value recommendation model based on the first historical transaction data and the first transaction fluency of the first virtual resource includes: Predicting sample value information of the first virtual resource based on the first historical transaction data through a learning model to be trained; Determining a sample transaction fluency of the first virtual resource based on the sample value information and the first historical transaction data; Training the learning model based on the sample value information, the true value information, the sample transaction fluency and the first transaction fluency of the first virtual resource to obtain a value recommendation model.

4. The method according to claim 3, wherein The determining the sample transaction fluency of the first virtual resource based on the sample value information and the first historical transaction data includes: Determining at least one of a sample selling parameter and a sample transaction risk coefficient of the first virtual resource based on the sample value information and the first historical transaction data; Determining the sample transaction fluency of the first virtual resource based on at least one of the sample selling parameter and the sample transaction risk coefficient.

5. The method according to claim 3, characterized in that The training the learning model based on the sample value information, the true value information, the sample transaction fluency and the first transaction fluency of the first historical transaction data to obtain a value recommendation model includes: Determining a first loss function value based on the sample value information and the true value information; Determining a second loss function value based on the sample transaction fluency and the first transaction fluency of the first virtual resource; Training the learning model based on the first loss function value and the second loss function value to obtain a value recommendation model.

6. The method according to any one of claims 1-5, characterized in that, The determining the first transaction fluency of the first virtual resource based on the first historical transaction data includes: Determine at least one of a first selling parameter and a first transaction risk coefficient of the first virtual resource based on the first historical transaction data; Determine a first transaction fluency of the first virtual resource based on at least one of the first selling parameter and the first transaction risk coefficient.

7. The method according to claim 6, characterized in that Determining the first transaction fluency of the first virtual resource based on the first selling parameter and the first transaction risk coefficient includes: Performing a weighted summation process on the first selling parameter and the first transaction risk coefficient to obtain the first transaction fluency of the first virtual resource.

8. The method according to claim 6, characterized in that, The first selling parameter includes a first selling time conversion parameter. Determining the first selling parameter of the first virtual resource based on the first historical transaction data includes: Determine the average selling time of the first virtual resource based on the first historical transaction data; Determine the first selling time conversion parameter based on the average selling time and a time decay factor.

9. The method according to claim 6, wherein The first selling parameter includes a first transaction risk coefficient. Determining the first transaction risk coefficient of the first virtual resource based on the first historical transaction data includes: Determine the number of failed transactions and the number of successful transactions based on the first historical transaction data; Determine the first transaction risk coefficient of the first virtual resource based on the number of failed transactions and the number of successful transactions.

10. An information processing apparatus for virtual resources, characterized in that, The apparatus includes: A first acquisition module, configured to acquire first historical transaction data of a first virtual resource, where the first historical transaction data includes attribute information and true value information of the first virtual resource; A first determination module, configured to determine a first transaction fluency of the first virtual resource based on the first historical transaction data, where the first transaction fluency is used to indicate the fluency degree of the transaction of the first virtual resource; A model training module, configured to train a value recommendation model based on the first historical transaction data and the first transaction fluency of the first virtual resource; A second determination module, configured to determine predicted value information of a second virtual resource through the value recommendation model; A second acquisition module, configured to acquire second historical transaction data of trading the second virtual resource based on the predicted value information; A third determination module, configured to determine a second transaction fluency of the second virtual resource based on the second historical transaction data; A model update module, configured to update the value recommendation model based on the second transaction fluency, so as to determine value information of a virtual resource to be traded based on the updated value recommendation model.

11. An electronic device, characterized in that, It includes a processor and a memory, and the memory stores multiple instructions; the processor loads the instructions from the memory to execute the method according to any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores multiple instructions, and the instructions are suitable for being loaded by a processor to execute the method according to any one of claims 1 to 9.

Images