Data processing method and device, equipment and storage medium

Abstract

The embodiment of the invention relates to a data processing method and device, equipment and a storage medium. The method comprises the following steps: acquiring a reference data item, wherein the reference data item indicates a reference entity to perform conversion on a reference object within a preset time period; determining entity identification information of the target entity based on one or more candidate entities associated with the entity constraint condition; determining object identification information of the target object based on one or more candidate objects associated with the object constraint condition; and based on the entity identification information and the object identification information, generating a target data item for the reference data item as a part of training data of the object recommendation model, the target data item indicating that the target entity does not perform conversion on the target object. Therefore, the quality of the generated training data can be improved, and the recommendation ability of the object recommendation model and the quality of the object recommendation service can be improved.

Description

Technical Field

[0001] Example embodiments of the present disclosure generally relate to the field of computers, and more particularly, to methods, apparatuses, devices, and computer-readable storage media for data processing. Background Art

[0002] With the rapid development of Internet technology, more and more applications and websites are designed to have functions related to object recommendation. Such applications and websites can provide object recommendation services with the help of object recommendation models. The recommendation ability of object recommendation models will be affected by the sample data sets used to train the models, that is, the quality of object recommendation services will be affected by the sample data sets. Therefore, people expect to generate higher quality sample data sets, so as to improve the recommendation ability of object recommendation models and the quality of object recommendation services. Summary of the invention

[0003] In a first aspect of the present disclosure, a method for data processing is provided. The method includes: obtaining a reference data item, the reference data item indicating that a reference entity performs a transformation on a reference object within a predetermined time period; determining entity identification information of a target entity based on one or more candidate entities associated with an entity constraint; determining object identification information of the target object based on one or more candidate objects associated with an object constraint; and generating a target data item for the reference data item as part of training data for an object recommendation model based on the entity identification information and the object identification information, the target data item indicating that the target entity does not perform a transformation on the target object.

[0004] In a second aspect of the present disclosure, a device for data processing is provided. The device includes: a data item acquisition module configured to acquire a reference data item, the reference data item indicating that the reference entity performs a transformation on the reference object within a predetermined time period; an entity identification determination module configured to determine the entity identification information of the target entity based on one or more candidate entities associated with the entity constraint; an object identification determination module configured to determine the object identification information of the target object based on one or more candidate objects associated with the object constraint; and a data item generation module configured to generate a target data item for the reference data item as part of the training data of the object recommendation model based on the entity identification information and the object identification information, the target data item indicating that the target entity does not perform a transformation on the target object.

[0005] In a third aspect of the present disclosure, an electronic device is provided. The device includes at least one processing unit; and at least one memory, the at least one memory is coupled to the at least one processing unit and stores instructions for execution by the at least one processing unit. When the instructions are executed by the at least one processing unit, the device executes the method of the first aspect.

[0006] In a fourth aspect of the present disclosure, a computer-readable storage medium is provided, wherein a computer program is stored on the computer-readable storage medium, and the computer program can be executed by a processor to implement the method of the first aspect.

[0007] It should be understood that the contents described in this content section are not intended to limit the key features or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, wherein:

[0009] Figure 1 A schematic diagram showing an example environment in which embodiments of the present disclosure can be implemented;

[0010] Figure 2 A flowchart showing a process for data processing according to some embodiments of the present disclosure is shown;

[0011] Figure 3 A schematic diagram showing an example of a target data item according to some embodiments of the present disclosure;

[0012] Figure 4 A schematic diagram showing an example process of data processing according to some embodiments of the present disclosure;

[0013] Figure 5 A schematic structural block diagram showing an apparatus for data processing according to some embodiments of the present disclosure; and

[0014] Figure 6 A block diagram of an electronic device is shown in which one or more embodiments of the present disclosure may be implemented. DETAILED DESCRIPTION

[0015] It is understandable that before using the technical solutions disclosed in the embodiments of the present disclosure, the types, scope of use, usage scenarios, etc. of the personal information involved in the present disclosure should be informed to the user and the user's authorization should be obtained in an appropriate manner in accordance with relevant laws and regulations.

[0016] For example, in response to receiving an active request from a user, a prompt message is sent to the user to clearly prompt the user that the operation requested to be performed will require obtaining and using the user's personal information. Thus, the user can autonomously choose whether to provide personal information to software or hardware such as an electronic device, application, server, or storage medium that performs the operation of the technical solution of the present disclosure according to the prompt message.

[0017] As an optional but non-limiting implementation, in response to receiving an active request from the user, the prompt information may be sent to the user in the form of a pop-up window, in which the prompt information may be presented in text form. In addition, the pop-up window may also carry a selection control for the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0018] It is understandable that the above notification and the process of obtaining user authorization are merely illustrative and do not constitute a limitation on the implementation of the present disclosure. Other methods that meet the relevant laws and regulations may also be applied to the implementation of the present disclosure.

[0019] It is understandable that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and relevant provisions.

[0020] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments set forth herein. On the contrary, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0021] In the description of the embodiments of the present disclosure, the term "including" and similar terms should be understood as open inclusion, that is, "including but not limited to". The term "based on" should be understood as "based at least in part on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The term "some embodiments" should be understood as "at least some embodiments". Other explicit and implicit definitions may be included below. The terms "first", "second", etc. may refer to different or the same objects. Other explicit and implicit definitions may be included below.

[0022] As used herein, the term "model" can learn the association between the corresponding input and output from the training data, so that after the training is completed, the corresponding output can be generated for a given input. The generation of the model can be based on machine learning technology. Deep learning is a machine learning algorithm that processes inputs and provides corresponding outputs by using multiple layers of processing units. In this article, "model" may also be referred to as "machine learning model", "machine learning network" or "network", which are used interchangeably in this article. A model may also include different types of processing units or networks.

[0023] Generally, machine learning can be roughly divided into three stages, namely the training stage, the testing stage, and the application stage (also called the inference stage). In the training stage, a given model can be trained using a large amount of training data, and the parameter values ​​are continuously updated iteratively until the model can obtain consistent inferences that meet the expected goals from the training data. Through training, the model can be considered to be able to learn the association from input to output (also called the mapping of input to output) from the training data. The parameter values ​​of the trained model are determined. In the testing stage, the test input is applied to the trained model to test whether the model can provide the correct output, thereby determining the performance of the model. In the application stage, the model can be used to process the actual input based on the parameter values ​​obtained from the training to determine the corresponding output.

[0024] Embodiments of the present disclosure relate to recommending objects to entities. In some cases, an entity may perform a conversion or transformation operation on an object. The entity here may include, but is not limited to, an individual, an organization, an institution, a user, etc., and the object may include, but is not limited to, an item, media content (e.g., a video), etc. The conversion is a resource demand behavior that an entity may perform on an object, which may include, but is not limited to, purchase, click, browse, register, download, etc.

[0025] As briefly mentioned above, object recommendation services can be provided with the help of object recommendation models. The recommendation ability of the object recommendation model will be affected by the sample dataset used to train the model, that is, the quality of the object recommendation service will be affected by the sample dataset. Therefore, it is expected that a higher quality sample dataset can be generated to improve the recommendation ability of the object recommendation model.

[0026] Each data item (also referred to as a sample) in the sample data set may include at least three key attributes: entity identification information, object identification information, and a sample label indicating whether the entity performs a transformation on the object. The data item corresponding to the sample label indicating that the entity performs a transformation on the object may also be referred to as a positive feedback data item or a positive example, and the data item corresponding to the sample label indicating that the entity does not perform a transformation on the object may also be referred to as a negative feedback data item or a negative example. For example, the data item "entity A-object A-performs transformation" is a positive example, and the data item "entity A-object A-does not perform transformation" is a negative example.

[0027] Generally speaking, a sample data set with a balanced number of positive feedback data items and negative feedback data items is used to train the model. In some cases, only positive feedback data items may be available, for example, in a scenario of shopping in a physical store. Therefore, in this case, traditionally, only positive feedback data items are often used to generate a sample data set, and this sample data set including only positive feedback data is used to train the object recommendation model. Since the number of positive feedback data items is often less than that of negative feedback data items, the recommendation effect of the object recommendation model trained in this way is often poor for entities and / or objects that include less positive feedback data.

[0028] To this end, an embodiment of the present disclosure proposes an improved scheme for data processing. According to various embodiments of the present disclosure, a reference data item indicating that a reference entity performs a transformation on a reference object within a predetermined time period is obtained. Based on one or more candidate entities associated with entity constraints, entity identification information of a target entity is determined. Based on one or more candidate objects associated with object constraints, object identification information of the target object is determined. Based on the entity identification information and the object identification information, a target data item is generated for the reference data item. The target data item serves as part of the training data of an object recommendation model. The target data item indicates that the target entity does not perform a transformation on the target object.

[0029] Thus, based on the reference data item (i.e., positive example) indicating that the reference entity has performed a transformation on the reference object within a predetermined time period, a target data item (i.e., negative example) indicating that the target entity has not performed a transformation on the target object can be generated. The generated target data item will be used as part of the training data to train the object recommendation model. In this way, the quality of the generated training data can be improved, which helps to improve the recommendation ability of the object recommendation model and improve the quality of the object recommendation service.

[0030] Figure 1 1 is a schematic diagram of an example environment 100 in which embodiments of the present disclosure can be implemented. Figure 1 As shown, example environment 100 may include electronic device 110 .

[0031] In the example environment 100, the electronic device 110 can obtain the reference data item 102. It can be understood that although Figure 1 Only one reference data item is shown in the figure, but the reference data item 102 may include multiple reference data items, and the present disclosure does not limit the number of reference data items. For each reference data item 102, the electronic device 110 may generate multiple target data items 104 based on the reference data item 102 (for example, it may include target data items 104-1, 104-2, ..., 104-N, where N is a positive integer greater than or equal to 1).

[0032] The electronic device 110 can generate training data for training the object recommendation model 115 based on the reference data item 102 and the generated target data item 104. Such training data can also be referred to as a sample data set, sample data, sample set, etc. The object recommendation model 115 can be, for example, any neural network that can perform object recommendation, including but not limited to a fully convolutional network (FCN), a convolutional neural network (CNN), a recurrent neural network (RNN), etc., and the embodiments of the present disclosure are not limited in this respect. In some embodiments, the object recommendation model 115 can be stored locally on the electronic device 110, and the electronic device 110 can directly use the local object recommendation model 115 to implement object recommendation. In some embodiments, the object recommendation model 115 can also be a model stored in the cloud, and the electronic device 110 can call the object recommendation model 115 stored in the cloud to implement object recommendation when it needs to perform related tasks.

[0033] The electronic device 110 may include any computing system with computing capabilities, such as various computing devices / systems, terminal devices, server devices, etc. The terminal device may be any type of mobile terminal, fixed terminal or portable terminal, etc., including a mobile phone, a desktop computer, a laptop computer, a notebook computer, a netbook computer, a tablet computer, a media computer, a multimedia tablet, a personal communication system (PCS) device, a personal navigation device, a personal digital assistant (PDA), an audio / video player, a digital camera / camcorder, a positioning device, a television receiver, a radio broadcast receiver, an e-book device, a virtual reality (VR) all-in-one machine, a game console, a game book, or any combination of the foregoing, including accessories and peripherals of these devices or any combination thereof.

[0034] The server-side device can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or 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 distribution networks, and big data and artificial intelligence platforms. Server-side devices can include computing systems / servers such as mainframes, edge computing nodes, computing devices in cloud environments, and so on.

[0035] Although some example application scenarios of interest tag sets are discussed above, it should be understood that this is only one of many application scenarios of the present disclosure. The embodiments of the present disclosure are also applicable to other application scenarios. Therefore, the embodiments of the present disclosure are not limited in this respect.

[0036] It should also be understood that the environment 100 and its structure and functionality are described for exemplary purposes only and do not imply any limitation on the scope of the present disclosure.

[0037] Figure 2 1 shows a flowchart of a process 200 for data processing according to some embodiments of the present disclosure. The process 200 may be implemented at the electronic device 110. For ease of discussion, reference will be made to Figure 1 The process 200 is described based on the environment 100. It should be noted that the operations performed by the electronic device 110 described above and the operations performed by the electronic device 110 described later may be performed by related applications installed on the electronic device 110.

[0038] In block 210, the electronic device 110 obtains a reference data item, which indicates that the reference entity performs a conversion on the reference object within a predetermined time period. The predetermined time period here may be any appropriate time period, which may be, for example, 1 month, 15 days, 10 days, 1 week, etc., and the present disclosure does not limit the specific time period. The reference data item may be, for example, a positive example or a positive feedback data item.

[0039] The electronic device 110 may obtain the reference data item (i.e., the positive example) in any appropriate manner. For example, the electronic device 110 may obtain the reference data item uploaded by the user, obtain the reference data item sent by other electronic devices, obtain the reference data item from a server, etc. In some embodiments, the reference data item may also be generated by the electronic device 110 itself based on the acquired data.

[0040] At block 220 , the electronic device 110 determines entity identification information of the target entity based on one or more candidate entities associated with the entity constraint.

[0041] In some embodiments, the electronic device 110 may acquire a set of entity constraints for an entity group. This set of entity constraints may, for example, include a first entity constraint corresponding to a first distinction difficulty and a second entity constraint corresponding to a second distinction difficulty. The first distinction difficulty and the second distinction difficulty here respectively indicate the difficulty of the object recommendation model in predicting whether an entity performs an object conversion, and the second distinction difficulty is greater than the first distinction difficulty. In some embodiments, the first entity constraint with a lower corresponding difficulty may be referred to as a simple constraint, a simple sampling condition, etc., and the second entity constraint with a higher corresponding difficulty may be referred to as a difficult constraint, a difficult sampling condition, etc.

[0042] In some embodiments, the first entity constraint condition may include, for example, that the entity has not performed object conversion within a predetermined time period, that the entity belongs to a candidate entity set, etc. The second entity constraint condition may include, for example, that the entity has performed object conversion in history but has not performed object conversion within a predetermined time period, that the entity has performed object conversion in a predetermined length of historical time period before the predetermined time period but has not performed object conversion within the predetermined time period, and that the entity has performed object conversion within the predetermined time period, etc. It should be noted that if the second entity constraint condition includes that the entity has performed object conversion within a predetermined time period, since the reference data item indicates that the reference entity has performed conversion on the reference object within the predetermined time period, the candidate entity obtained based on such a second entity constraint condition is the same entity as the reference entity.

[0043] In some embodiments, the electronic device 110 may select entity constraints from a set of entity constraints for an entity group. The electronic device 110 may then select a subset of candidate entities that satisfy the entity constraints from the candidate entity set. The candidate entity set includes a large number of candidate entities. The electronic device 110 may randomly select a certain number of candidate entities from the candidate entity set based on the selected entity constraints, and these candidate entities constitute a subset of candidate entities that satisfy the entity constraints. The number of candidate entities in the candidate entity subset here may be pre-set, or determined by the electronic device 110 itself, and the present disclosure does not limit the specific number and the method for determining the number.

[0044] In some embodiments, the electronic device 110 may randomly select an entity constraint from a set of entity constraints. For example, the electronic device 110 may randomly select a first entity constraint from a first entity constraint and a second entity constraint. The electronic device 110 then selects a candidate entity subset that satisfies the entity constraint from the candidate entity set based on the first entity constraint. That is, if the candidate entity subset includes 10 candidate entities, all of the 10 candidate entities satisfy the first entity constraint.

[0045] In some embodiments, the electronic device 110 may also select entity constraints according to certain rules. This rule may, for example, indicate the probability of selecting different entity constraints. For example, the electronic device 110 may determine that the probability of selecting the first entity constraint is 20% and the probability of selecting the second entity constraint is 80% according to the rule. In this case, if a candidate entity subset including 10 candidate entities is to be determined, the electronic device 110 may, for example, determine 2 candidate entities based on the first entity constraint, determine 8 candidate entities based on the second entity constraint, and then determine the candidate entity subset based on the determined 10 candidate entities.

[0046] The electronic device 110 may generate entity identification information based on identification information of candidate entities in the candidate entity subset. It can be understood that the entity identification information includes identification information of at least one candidate entity in the candidate entity subset.

[0047] At block 230 , the electronic device 110 determines object identification information of the target object based on the one or more candidate objects associated with the object constraint.

[0048] In some embodiments, the electronic device 110 can obtain a set of object constraints for an object group. This set of object constraints may, for example, include a first object constraint corresponding to a third distinction difficulty and a second object constraint corresponding to a fourth distinction difficulty. The third distinction difficulty and the fourth distinction difficulty here respectively indicate the difficulty of the object recommendation model predicting whether an object is converted into an object (for example, the difficulty of whether an item is purchased), and the fourth distinction difficulty is greater than the third distinction difficulty. Similarly, the third entity constraint with a lower corresponding difficulty can be referred to as a simple constraint, a simple sampling condition, etc., and the fourth entity constraint with a higher corresponding difficulty can be referred to as a difficult constraint, a difficult sampling condition, etc.

[0049] In some embodiments, the first object constraint condition may include, for example, that the object belongs to a set of candidate objects. The second object constraint condition may include, for example, that the historical conversion rate of the object is higher than a threshold value (for example, the purchase rate of the object is higher than a purchase rate threshold value), that the object is of the same type as the reference object (that is, the object is an object of the same type as the reference object), that the object and the reference object are provided by the same provider (for example, the object and the reference object are objects of the same brand), that the object and the reference object have matching conversion costs (for example, the object and the reference object have the same price), and the like. It should be noted that, when the second object constraint condition is selected, the electronic device 110 needs to determine the candidate object based on the reference object, that is, the candidate object determined by the electronic device 110 is determined for the reference object. In this case, the target data item generated based on the identification information of such a candidate object is a data item generated for the reference object.

[0050] This set of object constraints may also include, for example, a third object constraint related to visibility of the object to the entity (also referred to as a business constraint). For example, the third object constraint may indicate that the object is an object that the entity is likely to have seen (e.g., a main promoted object in a consumption scenario).

[0051] Similar to selecting a subset of candidate entities that satisfy entity constraints, the electronic device 110 may select object constraints from a set of object constraints for an object group. The electronic device 110 may then select a subset of candidate objects that satisfy object constraints from the candidate object set. The candidate object set includes a large number of candidate objects. The electronic device 110 may randomly select a certain number of candidate objects from the candidate object set based on the selected object constraints, and these candidate objects constitute a subset of candidate objects that satisfy object constraints. Similarly, the number of candidate objects in the candidate object subset here may be pre-set, or determined by the electronic device 110 itself, and the present disclosure does not limit the specific number and the method for determining the number.

[0052] In some embodiments, the electronic device 110 may randomly select an object constraint from a set of object constraints, and then the electronic device 110 selects a certain number of candidate objects that meet the selected object constraint from the candidate object set based on the selected object constraint, and these candidate objects constitute a candidate object subset.

[0053] In some embodiments, the electronic device 110 may also select object constraints according to certain rules. This rule may, for example, indicate the probability of selecting different object constraints. For example, the electronic device 110 may determine, according to the rule, that the probability of selecting the first object constraint is 30%, the probability of selecting the second object constraint is 50%, and the probability of selecting the third object constraint is 20%. In this case, to determine a candidate object subset including 10 candidate objects, the electronic device 110 may, for example, determine 3 candidate objects based on the first object constraint, determine 5 candidate objects based on the second object constraint, determine 2 candidate objects based on the third object constraint, and then determine the candidate object subset based on the determined 10 candidate objects.

[0054] The electronic device 110 may generate object identification information based on identification information of candidate objects in the candidate object subset. It can be understood that the object identification information includes identification information of at least one candidate object in the candidate object subset.

[0055] In block 240 , the electronic device 110 generates a target data item for the reference data item as a part of training data for the object recommendation model based on the entity identification information and the object identification information, the target data item indicating that the target entity has not performed a conversion on the target object.

[0056] In some embodiments, the electronic device 110 may generate a plurality of target data items (ie, negative examples) for the reference data items based on the acquired entity identification information and object identification information. Figure 3A schematic diagram 300 is shown of an example of a target data item according to some embodiments of the present disclosure. Figure 3 As shown, the electronic device 110 can obtain entity identification information 310 and object identification information 320. The entity identification information 310 includes identification information of a simple sampling entity 312 obtained based on a simple constraint condition for the entity (i.e., the first entity constraint condition) and identification information of a difficult sampling entity 314 obtained based on a difficult constraint condition for the entity (i.e., the second constraint condition). It can be understood that both the simple sampling entity 312 and the difficult sampling entity 314 can include multiple candidate entities. Similarly, the object identification information 320 includes identification information of a simple sampling object 322 obtained based on a simple constraint condition for the object (i.e., the first object constraint condition), identification information of a difficult sampling object 324 obtained based on a difficult constraint condition for the object (i.e., the second object constraint condition), and identification information of a business sampling object 326 obtained based on a business constraint condition (i.e., the third object constraint condition). Similarly, each of the simple sampling object 322, the difficult sampling object 324, and the business sampling object 326 can include multiple candidate objects.

[0057] For example, the electronic device 110 may randomly combine the identification information in the entity identification information 310 with the identification information in the object identification information 320 in pairs, and add a label indicating that the conversion is not performed to the combination result to generate multiple target data items 104. For example, the electronic device 110 may combine the identification information of a simple sampling entity 312 (for example, it may be simply referred to as entity A) and the identification information of a simple sampling object 322 (for example, it may be simply referred to as object a), and add a label indicating that the conversion is not performed to the combination to generate the target data item "entity A-object a-no conversion". It can be understood that the electronic device 110 can generate multiple target data items 104.

[0058] Since the first object constraint condition includes that the object belongs to a candidate object set, there is a situation where the simple sampling object 322 is the same as the reference object. Since the first entity constraint condition includes that the entity belongs to a candidate entity set and the second entity constraint condition includes that the entity has performed object transformation within a predetermined time period, there is a situation where the simple sampling entity 312 is the same as the reference entity. In some embodiments, in order to improve the accuracy of the generated target data items 104, the electronic device 110 can also screen the generated multiple target data items 104 to avoid the situation where the target object and target entity in the target data item 104 are the same as the reference object and reference entity in the reference data item 102. It can be understood that the electronic device 110 can retain a target data item that is the same as the reference object / reference entity among the target objects and target entities. Exemplarily, if the generated multiple target data items 104 include the data item "entity A-object A-no conversion", the data item "entity A-object B-no conversion" and the data item "entity C-object A-no conversion", and the reference data item 102 is "entity A-object A-no conversion", the electronic device 110 can filter out the data item "entity A-object A-no conversion" and retain the data item "entity A-object B-no conversion" and the data item "entity C-object A-no conversion".

[0059] In some embodiments, when the reference data item 102 includes multiple reference data items, the electronic device 110 may generate a target data item for each reference data item in sequence. Figure 4 A schematic diagram of an example process 400 of data processing according to some embodiments of the present disclosure is shown. Process 400 can be considered as an example implementation of process 200.

[0060] In block 401 , the electronic device 110 iterates through each reference data item.

[0061] For the current reference data item, the electronic device 110 selects an entity constraint in block 402. The electronic device 110 may select the entity constraint according to a rule indicating the probability of selecting different entity constraints, for example.

[0062] In block 403, the electronic device 110 determines entity identification information based at least on the selected entity constraint. In some embodiments, when the selected entity constraint is a second entity constraint, and the second entity constraint includes that the entity has performed object transformation within a predetermined time period, the electronic device 110 may determine the entity identification information based on the reference data item and the entity constraint.

[0063] At block 404, the electronic device 110 selects an object constraint. The electronic device 110 may select the object constraint, for example, according to a rule indicating the probability of selecting different object constraints.

[0064] At block 405, the electronic device 110 determines object identification information based at least on the object constraint. In some embodiments, when the selected object constraint is the second object constraint, the electronic device 110 may determine the object identification information based on the reference data item and the object constraint.

[0065] In block 406 , the electronic device 110 generates a target data item based on the object identification information and the entity identification information.

[0066] In block 407, the electronic device 110 determines whether the number of target data items has reached the target. For example, if the target indicates that 10 target data items should be generated for each reference data item, the electronic device 110 may determine whether the number of target data items currently generated has reached 10. If the number of target data items has not reached the target, the electronic device 110 may continue to generate target data items for the current reference data item according to the method shown in blocks 402 to 406.

[0067] When the number of target data items reaches the target, the electronic device 110 may process the next reference data item, that is, the electronic device 110 may generate a target data item of target data for the next reference data item in block 408. The electronic device 110 may perform process 400 until a target data item that meets the target number is generated for each reference data item.

[0068] Return to reference Figure 3 , the electronic device 110 can construct a sample set 305 based on the reference data item 102 and the generated target data item 104. The sample set 305 is training data for training the object recommendation model.

[0069] In summary, according to various embodiments of the present disclosure, a target data item indicating that a target entity has not performed a transformation on a target object can be generated based on a reference data item indicating that a reference entity has performed a transformation on a reference object within a predetermined time period. The generated target data item will be used as part of the training data to train the object recommendation model. In this way, the quality of the generated training data can be improved, which helps to improve the recommendation capability of the object recommendation model and improve the quality of the object recommendation service.

[0070] The embodiments of the present disclosure also provide corresponding devices for implementing the above methods or processes. Figure 5A schematic structural block diagram of an apparatus 500 for data processing according to some embodiments of the present disclosure is shown. The apparatus 500 may be implemented as or included in the electronic device 110. Each module / component in the apparatus 500 may be implemented by hardware, software, firmware or any combination thereof.

[0071] As shown in the figure, the device 500 includes a data item acquisition module 510, which is configured to acquire a reference data item, and the reference data item indicates that the reference entity performs a transformation on the reference object within a predetermined time period. The device 500 also includes an entity identification determination module 520, which is configured to determine the entity identification information of the target entity based on one or more candidate entities associated with the entity constraint condition. The device 500 also includes an object identification determination module 530, which is configured to determine the object identification information of the target object based on one or more candidate objects associated with the object constraint condition. The device 500 also includes a data item generation module 540, which is configured to generate a target data item for the reference data item as part of the training data of the object recommendation model based on the entity identification information and the object identification information, and the target data item indicates that the target entity does not perform a transformation on the target object.

[0072] In some embodiments, the entity identification determination module 520 is further configured to: select entity constraints from a set of entity constraints for an entity group; select a subset of candidate entities that satisfy the entity constraints from a set of candidate entities; and generate entity identification information based on identification information of candidate entities in the subset of candidate entities.

[0073] In some embodiments, a set of entity constraints includes: a first entity constraint corresponding to a first distinction difficulty; a second entity constraint corresponding to a second distinction difficulty, wherein the first distinction difficulty and the second distinction difficulty respectively indicate the difficulty of the object recommendation model predicting whether an entity performs object transformation, and the second distinction difficulty is greater than the first distinction difficulty.

[0074] In some embodiments, the first entity constraint includes at least one of the following: the entity has not performed object conversion within a predetermined time period, or the entity belongs to a candidate entity set, and the second entity constraint includes at least one of the following: the entity has historically performed object conversion but has not performed object conversion within the predetermined time period, the entity has performed object conversion in a historical time period of a predetermined length before the predetermined time period but has not performed object conversion within the predetermined time period, and the entity has performed object conversion within the predetermined time period.

[0075] In some embodiments, the object identification determination module 530 is further configured to: select object constraints from a set of object constraints for an object group; select a subset of candidate objects that satisfy the object constraints from a set of candidate objects; and generate object identification information based on identification information of the candidate objects in the subset of candidate objects.

[0076] In some embodiments, a set of object constraints includes: a first object constraint corresponding to a third distinction difficulty; a second object constraint corresponding to a fourth distinction difficulty, wherein the third distinction difficulty and the fourth distinction difficulty respectively indicate the difficulty of the object recommendation model predicting whether an object is to be transformed into an executed object, and the fourth distinction difficulty is greater than the third distinction difficulty.

[0077] In some embodiments, the first object constraint includes: the object belongs to a set of candidate objects, and the second object constraint includes at least one of the following: the object's historical conversion rate is higher than a threshold, the object and the reference object are of the same type, the object and the reference object are provided by the same provider, or the object and the reference object have matching conversion costs.

[0078] In some embodiments, the set of object constraints further includes: a third object constraint related to visibility of the object to the entity.

[0079] In some embodiments, the target data item is one of a plurality of data items generated for a reference object.

[0080] The modules and / or units included in the device 500 can be implemented in various ways, including software, hardware, firmware, or any combination thereof. In some embodiments, one or more modules and / or units can be implemented using software and / or firmware, such as machine executable instructions stored on a storage medium. In addition to or as an alternative to machine executable instructions, some or all of the modules and / or units in the device 500 can be implemented at least in part by one or more hardware logic components. As an example and not limitation, exemplary types of hardware logic components that can be used include field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0081] Figure 6 1 shows a block diagram of an electronic device 600 in which one or more embodiments of the present disclosure may be implemented. It should be understood that Figure 6 The electronic device 600 shown is merely exemplary and should not constitute any limitation on the functionality and scope of the embodiments described herein. Figure 6 The electronic device 600 shown can be used to implement Figure 1 An electronic device 110.

[0082] like Figure 6 As shown, the electronic device 600 is in the form of a general electronic device. The components of the electronic device 600 may include, but are not limited to, one or more processors or processing units 610, a memory 620, a storage device 630, one or more communication units 640, one or more input devices 650, and one or more output devices 660. The processing unit 610 may be an actual or virtual processor and is capable of performing various processes according to a program stored in the memory 620. In a multi-processor system, multiple processing units execute computer executable instructions in parallel to improve the parallel processing capability of the electronic device 600.

[0083] The electronic device 600 typically includes a plurality of computer storage media. Such media may be any accessible media that is accessible to the electronic device 600, including but not limited to volatile and non-volatile media, removable and non-removable media. The memory 620 may be a volatile memory (e.g., registers, caches, random access memory (RAM)), a non-volatile memory (e.g., a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), flash memory), or some combination thereof. The storage device 630 may be a removable or non-removable medium, and may include a machine-readable medium, such as a flash drive, a disk, or any other medium, which may be capable of being used to store information and / or data and may be accessed within the electronic device 600.

[0084] The electronic device 600 may further include additional removable / non-removable, volatile / non-volatile storage media. Figure 6 As shown in , a disk drive for reading or writing from a removable, non-volatile disk (e.g., a "floppy disk") and an optical drive for reading or writing from a removable, non-volatile optical disk may be provided. In these cases, each drive may be connected to the bus (not shown) by one or more data media interfaces. The memory 620 may include a computer program product 625 having one or more program modules that are configured to perform various methods or actions of various embodiments of the present disclosure.

[0085] The communication unit 640 implements communication with other electronic devices through a communication medium. Additionally, the functions of the components of the electronic device 600 can be implemented with a single computing cluster or multiple computing machines that can communicate through a communication connection. Therefore, the electronic device 600 can operate in a networked environment using a logical connection with one or more other servers, a network personal computer (PC), or another network node.

[0086] The input device 650 may be one or more input devices, such as a mouse, a keyboard, a tracking ball, etc. The output device 660 may be one or more output devices, such as a display, a speaker, a printer, etc. The electronic device 600 may also communicate with one or more external devices (not shown) through the communication unit 640 as needed, such as a storage device, a display device, etc., communicate with one or more devices that allow a user to interact with the electronic device 600, or communicate with any device that allows the electronic device 600 to communicate with one or more other electronic devices (e.g., a network card, a modem, etc.). Such communication may be performed via an input / output (I / O) interface (not shown).

[0087] According to an exemplary implementation of the present disclosure, a computer-readable storage medium is provided, on which computer-executable instructions are stored, wherein the computer-executable instructions are executed by a processor to implement the method described above. According to an exemplary implementation of the present disclosure, a computer program product is also provided, which is tangibly stored on a non-transitory computer-readable medium and includes computer-executable instructions, and the computer-executable instructions are executed by a processor to implement the method described above.

[0088] Various aspects of the present disclosure are described herein with reference to the flowcharts and / or block diagrams of the methods, devices, equipment, and computer program products implemented according to the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram can be implemented by computer-readable program instructions.

[0089] These computer-readable program instructions can be provided to a processing unit of a general-purpose computer, a special-purpose computer, or other programmable data processing device, thereby producing a machine, so that when these instructions are executed by the processing unit of the computer or other programmable data processing device, a device that implements the functions / actions specified in one or more boxes in the flowchart and / or block diagram is generated. These computer-readable program instructions can also be stored in a computer-readable storage medium, and these instructions cause the computer, programmable data processing device, and / or other equipment to work in a specific manner, so that the computer-readable medium storing the instructions includes a manufactured product, which includes instructions for implementing various aspects of the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0090] Computer-readable program instructions can be loaded onto a computer, other programmable data processing apparatus, or other device so that a series of operational steps are performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, so that the instructions executed on the computer, other programmable data processing apparatus, or other device implement the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0091] The flow chart and block diagram in the accompanying drawings show the possible architecture, function and operation of the system, method and computer program product according to multiple implementations of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a part of a module, program segment or instruction, and a part of a module, program segment or instruction includes one or more executable instructions for realizing the logical function of the specification. In some implementations as replacements, the function marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two continuous square boxes can actually be executed substantially in parallel, and they can sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be realized by a special hardware-based system that performs the function or action of the specification, or can be realized by a combination of special hardware and computer instructions.

[0092] The above descriptions of various implementations of the present disclosure are exemplary, non-exhaustive, and not limited to the disclosed implementations. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described implementations. The selection of terms used herein is intended to best explain the principles of the implementations, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the various implementations disclosed herein.

Claims

1. A data processing method, comprising: obtaining a reference data item, the reference data item indicating that a reference entity performs a transformation on a reference object within a predetermined time period; Determining entity identification information of a target entity based on one or more candidate entities associated with the entity constraint; Determining object identification information of a target object based on one or more candidate objects associated with the object constraint; as well as Based on the entity identification information and the object identification information, a target data item for the reference data item is generated as a part of training data of an object recommendation model, the target data item indicating that the target entity has not performed a conversion on the target object.

2. The method according to claim 1, wherein determining the entity identification information of the target entity comprises: selecting the entity constraint from a set of entity constraints for a group of entities; Selecting a subset of candidate entities that satisfy the entity constraint from the candidate entity set; as well as The entity identification information is generated based on identification information of candidate entities in the candidate entity subset.

3. The method of claim 2, wherein the set of entity constraints comprises: A first entity constraint corresponding to a first differentiation difficulty; The second entity constraint corresponding to the second differentiation difficulty, The first distinction difficulty and the second distinction difficulty respectively indicate the difficulty of the object recommendation model predicting whether an entity performs object transformation, and the second distinction difficulty is greater than the first distinction difficulty.

4. The method according to claim 2, wherein The first entity constraint condition includes at least one of the following: The entity does not perform an object conversion within the predetermined time period, or The entity belongs to the candidate entity set, and The second entity constraint condition includes at least one of the following: The entity has historically performed object conversion, but has not performed object conversion within the predetermined time period, The entity has performed object conversion in a history time period of a predetermined length before the predetermined time period, but has not performed object conversion in the predetermined time period. The entity has performed object conversion within the predetermined time period.

5. The method according to claim 1, wherein determining the object identification information of the target object comprises: selecting the subject constraint from a set of subject constraints for a subject population; Selecting a subset of candidate objects that meet the object constraint from the candidate object set; as well as The object identification information is generated based on identification information of candidate objects in the candidate object subset.

6. The method of claim 5, wherein the set of object constraints comprises: A first object constraint corresponding to the third differentiation difficulty; The second object constraint corresponding to the fourth differentiation difficulty, The third distinction difficulty and the fourth distinction difficulty respectively indicate the difficulty of the object recommendation model predicting whether the object is to be converted into an executed object, and the fourth distinction difficulty is greater than the third distinction difficulty.

7. The method according to claim 5, wherein the first object constraint comprises: The object belongs to the candidate object set, and The second object constraint condition includes at least one of the following: The historical conversion rate of the object is higher than the threshold, object is of the same type as the reference object, The object is provided by the same provider as the reference object, or The object has a matching conversion cost with the reference object.

8. The method of claim 5, wherein the set of object constraints further comprises: A third object constraint related to the visibility of an object to an entity. 9 . The method according to claim 1 , wherein the target data item is one of a plurality of data items generated for the reference object.

10. A device for data processing, comprising: a data item acquisition module configured to acquire a reference data item, the reference data item indicating that a reference entity performs a transformation on a reference object within a predetermined time period; An entity identification determination module, configured to determine entity identification information of a target entity based on one or more candidate entities associated with the entity constraint condition; an object identification determination module configured to determine object identification information of a target object based on one or more candidate objects associated with the object constraint condition; as well as The data item generation module is configured to generate a target data item for the reference data item as part of the training data of the object recommendation model based on the entity identification information and the object identification information, wherein the target data item indicates that the target entity has not performed a conversion on the target object.

11. An electronic device, comprising: at least one processing unit; as well as At least one memory, the at least one memory being coupled to the at least one processing unit and storing instructions for execution by the at least one processing unit, the instructions causing the electronic device to perform the method according to any one of claims 1 to 9 when executed by the at least one processing unit.

12. A computer-readable storage medium having a computer program stored thereon, wherein the computer program can be executed by a processor to implement the method according to any one of claims 1 to 9.

Images