Methods, devices, equipment, media, and products for object recommendation

Abstract

The present disclosure provides a method and apparatus for object recommendation, an electronic device, a computer-readable storage medium, and a computer program product, relating to the field of computer technologies, and particularly to intelligent recommendation technologies. The implementation solution is as follows: in response to determining that user tags of a target user for multiple service scenarios are the same, based on the user characteristics of the target user and the object characteristics of candidate objects for each service scenario, determine the recommendation characteristics of the target user for the multiple service scenarios; based on the recommendation characteristics, determine the recommended objects of the target user for the multiple service scenarios, where the recommended objects are determined from the candidate objects for the multiple service scenarios.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technologies, and more particularly, to intelligent recommendation technologies, and specifically to a method for object recommendation, an apparatus therefor, an electronic device, a computer-readable storage medium, and a computer program product. Background Art

[0002] A recommendation system can select objects that a user may be interested in from a large amount of data and recommend the objects to the user. Currently, recommendation systems can be applied to many business scenarios such as product recommendation, advertisement placement, friend recommendation, and so on. With the continuous increase in the number of users and the continuous increase in business scenarios, the recommendation effect of the recommendation system may be affected, or the recommendation cost may increase. In some business scenarios, the recommendation system may not accurately recommend objects to users.

[0003] The methods described in this section are not necessarily methods that have been previously conceived or adopted. Unless otherwise specified, any method described in this section should not be considered to be prior art merely because it is included in this section. Similarly, unless otherwise specified, the problems mentioned in this section should not be considered to have been recognized in any prior art. Summary of the Invention

[0004] The present disclosure provides a method for object recommendation, an apparatus therefor, an electronic device, a computer-readable storage medium, and a computer program product.

[0005] According to one aspect of the present disclosure, there is provided a method for object recommendation, including: in response to determining that user tags of a target user for multiple business scenarios are the same, determining recommendation features of the target user for the multiple business scenarios based on user features of the target user and object features of candidate objects for each business scenario; and determining recommended objects of the target user for the multiple business scenarios based on the recommendation features, where the recommended objects are determined from candidate objects for the multiple business scenarios.

[0006] According to one aspect of the present disclosure, there is provided an apparatus for object recommendation, including: a feature splicing unit configured to, in response to determining that user tags of a target user for multiple business scenarios are the same, determine recommendation features of the target user for the multiple business scenarios based on user features of the target user and object features of candidate objects for each business scenario; and an object recommendation unit configured to determine recommended objects of the target user for the multiple business scenarios based on the recommendation features, where the recommended objects are determined from candidate objects for the multiple business scenarios.

[0007] According to one aspect of the present disclosure, there is provided an electronic device, including: a processor; and a memory storing a program, the program including instructions that, when executed by the processor, cause the processor to execute the above object recommendation method.

[0008] According to one aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium storing a computer program, the computer program including instructions that, when executed by a processor of a computing device, cause the computing device to execute the above object recommendation method.

[0009] According to one aspect of the present disclosure, there is provided a computer program product including a computer program, wherein the computer program, when executed by a processor, implements the steps of the above object recommendation method. Description of the Drawings

[0010] The drawings exemplarily illustrate embodiments and form a part of the specification, and are used together with the written description of the specification to explain the exemplary embodiments. The illustrated embodiments are for illustrative purposes only and do not limit the scope of the claims. In all the drawings, the same reference numerals refer to similar but not necessarily identical elements.

[0011] Figure 1 A flowchart showing an object recommendation method according to an exemplary embodiment of the present disclosure is shown;

[0012] Figure 2 A flowchart showing an object recommendation method according to an exemplary embodiment of the present disclosure is shown;

[0013] Figure 3 A flowchart showing an object recommendation method according to an exemplary embodiment of the present disclosure is shown;

[0014] Figure 4 A flowchart showing a method for training a recommendation model according to an exemplary embodiment of the present disclosure is shown;

[0015] Figure 5 A block diagram showing the structure of an object recommendation apparatus according to an exemplary embodiment of the present disclosure is shown;

[0016] Figure 6 A block diagram showing the structure of a recommendation model training apparatus according to an exemplary embodiment of the present disclosure is shown;

[0017] Figure 7 A block diagram showing an electronic device according to an embodiment of the present disclosure is shown. Detailed Description of the Embodiments

[0018] The following describes exemplary embodiments of the present disclosure with reference to the accompanying drawings. Various details of the embodiments of the present disclosure are included to facilitate understanding, and they should be considered merely exemplary. Therefore, those of ordinary skill in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope of the present disclosure. Similarly, descriptions of well-known functions and structures are omitted below for clarity and conciseness.

[0019] In the present disclosure, unless otherwise specified, the terms "first", "second", etc. are used to describe various elements and are not intended to limit the positional relationship, temporal relationship, or importance relationship of these elements. Such terms are only used to distinguish one element from another. In some examples, the first element and the second element may refer to the same instance of the element, and in certain cases, based on the context description, they may also refer to different instances.

[0020] In the description of various examples in the present disclosure, the terms used are only for the purpose of describing specific examples and are not intended to be limiting. Unless the context clearly indicates otherwise, if the number of elements is not specifically limited, the element may be one or more. In addition, the term "and / or" used in the present disclosure covers any one of the listed items and all possible combinations.

[0021] In the related art, the implementation of object recommendation is to recommend through user characteristics. The disadvantage of this implementation is that there is no pertinence when recommending in multiple business scenarios, resulting in low recommendation accuracy. It may also introduce user characteristics irrelevant to the business scenario or lose user characteristics related to the business scenario, affecting the recommendation effect.

[0022] To solve the above problems, the present disclosure provides a method for object recommendation. By screening different types of users according to different business scenarios, user characteristics and object characteristics of different business scenarios are fused in a targeted manner. At the same time, by recommending based on multiple goals, object recommendation based on the entire link is realized, improving the object recommendation effect.

[0023] The object recommendation method of the present disclosure can, for example, but not limited to, be applied to a material recommendation system. By adopting the method in the embodiments of the present disclosure, materials (such as coupons) can be recommended to users in a targeted manner in different business scenarios, realizing multi-goal optimization of the entire link (such as from receiving materials to using materials for purchase), improving the pertinence of the material recommendation system and the recommendation effect.

[0024] The following will describe the embodiments of the present disclosure in detail with reference to the accompanying drawings.

[0025] Figure 1 The flowchart of an object recommendation method 100 according to an exemplary embodiment of the present disclosure is shown. AsFigure 1 As shown, the method 100 for object recommendation includes: step S102, in response to determining that the user tags of the target user for multiple business scenarios are the same, based on the user characteristics of the target user and the object characteristics of the candidate objects for each business scenario, determining the recommendation characteristics of the target user for multiple business scenarios; step S104, based on the recommendation characteristics, determining the recommended objects of the target user for multiple business scenarios; wherein, the recommended objects are determined from the candidate objects for multiple business scenarios.

[0026] Thus, it is possible to integrate different business scenarios and user groups, fully learn the user's behavior habits in different business scenarios, and be able to recommend by specifically cross-referencing the characteristics of users and items / goods in different scenarios, so as to solve the problem of insufficient accuracy in multiple business scenarios.

[0027] In some embodiments, the candidate objects include at least one object, and each object can correspond to one or more business scenarios.

[0028] According to some embodiments, the user tag of the target user for each business scenario is determined based on the historical behavior information of the target user in this business scenario. The determined user tag can characterize the category of the historical behavior information of the target user in this business scenario.

[0029] Exemplarily, the user tags of the target user for each business scenario can include two types: the first user tag and the second user tag. It can be understood that the user tags of the target user for each business scenario can also include three or more types, which can be set according to specific business requirements and are not limited herein.

[0030] Exemplarily, the historical behavior information can include historical purchase information. For example, determining the user tags of the target user for each business scenario includes: in response to determining that the target user has no historical purchase information in this business scenario, determining the user tag as a new customer acquisition tag; and in response to determining that the target user has historical purchase information in this business scenario, determining the user tag as a repeat purchase tag. Thus, the target users for different business scenarios can be screened to make targeted recommendations.

[0031] In some embodiments, according to the frequency of the purchase behavior of the target user, the repeat purchase tag can also be determined as a high-frequency repeat purchase tag and a low-frequency repeat purchase tag. For example, when the frequency of the purchase behavior of the target user in this business scenario is greater than a certain predetermined threshold, the user tag of the target user for this business scenario can be set as a high-frequency repeat purchase tag. And when the frequency of the purchase behavior of the target user in this business scenario is less than or equal to the predetermined threshold, the user tag of the target user for this business scenario can be set as a low-frequency repeat purchase tag.

[0032] According to some embodiments, the method 100 for object recommendation further includes: determining candidate objects for each business scenario based on the user tags of the target user for each business scenario.

[0033] Thus, based on the division of user tags in different business scenarios, the candidate objects of the target user in that business scenario can be determined more accurately, thereby avoiding interference from irrelevant features on the recommendation.

[0034] Exemplarily, when the user tag of the target user in the first business scenario is the first user tag (such as the recruitment tag), at least one candidate object of the target user for the first business scenario is the first candidate object corresponding to the first user tag. When the user tag of another target user in the first business scenario is the second user tag (such as the repeat purchase tag), at least one candidate object of the other target user for the first business scenario is the second candidate object corresponding to the second user tag. In some business scenarios, the candidate objects corresponding to different user tags may be different. For example, in the above-mentioned first business scenario, the first candidate object may be different from the second candidate object. In some business scenarios, the candidate objects corresponding to different user tags may be at least partially the same. For example, in the above-mentioned first business scenario, some of the first candidate objects may be the same as the second candidate object. In certain business scenarios, the candidate objects corresponding to the recruitment tag may be the objects corresponding to the products that the target user has not purchased before, in order to recommend the target user to make the first purchase; while the candidate objects corresponding to the repeat purchase tag may be the objects corresponding to the products that the target user has already purchased, in order to recommend the target user to make a repeat purchase.

[0035] According to some embodiments, determining candidate objects for each business scenario based on the user tags of the target user for each business scenario further includes: determining the candidate objects of the target user for that business scenario based on the corresponding user tags and the historical behavior information of the target user in that business scenario.

[0036] Exemplarily, the historical behavior information may include whether the target user has received the object in the past. The candidate object of the target user for that business scenario is the object that has not been received by the target user in that business scenario. If the target user has received the object in the past, then the object is removed from at least one candidate object, so that at least one candidate object of the target user for that business scenario is the object that has not been received by the target user in that business scenario. This ensures that the target user will not be recommended the same object multiple times.

[0037] In this embodiment, the recommended objects for the user are determined from at least one candidate object in at least two business scenarios. The recommendation model can select the recommended objects from at least one candidate object. In some embodiments, the number of the recommended objects can be determined by user tags. For example, when the target user has a first user tag for multiple business scenarios, two recommended objects can be selected from the candidate objects in the multiple business scenarios. When the user tags of the target user for at least two business scenarios are the first user tag, and the user tags for other business scenarios are the second user tag, one recommended object is selected from the candidate objects in the at least two business scenarios, and another recommended object is selected from the candidate objects in the other business scenarios. The first user tag can be, for example, a recruitment tag, and the second user tag can be, for example, a repeat purchase tag.

[0038] In this embodiment, in step S104, based on the user characteristics of the target user and the object characteristics of at least one candidate object in each of at least two business scenarios, the recommendation characteristics of the target user for at least two business scenarios are determined, including: splicing the user characteristics of the target user and the object characteristics of the candidate object to obtain the recommendation characteristics of the target user for at least two business scenarios. In some embodiments, the user characteristics of the target user, the business scenario characteristics of at least two business scenarios, and the object characteristics of the candidate object can also be spliced to obtain the recommendation characteristics of the target user for at least two business scenarios.

[0039] Exemplarily, in some business scenarios, the user characteristics of the target user can be based on the basic information and behavior information of the target user. Among them, the basic information can include age, gender, occupation, etc.; the behavior information can include historical behavior information, such as purchase information, browsing information, click information, etc. The object characteristics of the candidate object can be based on the commodity information, object information, etc. corresponding to the candidate object. Among them, the commodity information can include the type, appearance, price, spokesperson, etc. of the commodity, and the object information can include the corresponding task information, reward information (such as coupon information), etc.

[0040] According to some embodiments, the object recommendation method 100 further includes: in response to determining that the target user does not have historical behavior information in each business scenario within a predetermined time range, directly determining the recommended objects for the target user based on a preset rule.

[0041] Exemplarily, the historical behavior information can be historical purchase information. That is, when the target user does not have historical purchase behavior in all business scenarios, objects are directly recommended for the target user based on a preset rule. The preset rule can be to directly recommend the object corresponding to the recruitment tag, or directly recommend a popular object with a relatively high purchase probability for the user.

[0042] Figure 2The flowchart of an object recommendation method according to an exemplary embodiment of the present disclosure is shown. As Figure 2 shown, and in combination with Figure 1 , step S104 includes: step S1042, based on the recommendation features, obtaining the predicted execution scores of multiple target tasks corresponding to the candidate objects, where the predicted execution scores are used to characterize the probability that the corresponding target tasks are executed by the target user; step S1044, based on the predicted execution scores of the multiple target tasks corresponding to each candidate object, determining the predicted recommendation scores corresponding to each candidate object; and step S1046, based on the predicted recommendation scores corresponding to each candidate object, determining the recommended objects for the target user for multiple business scenarios from the candidate objects.

[0043] In this embodiment, the predicted execution score is used to characterize the probability that the corresponding target task is executed by the target user. That is, when the prediction model determines that the target task has a high probability of being executed by the target user based on the recommendation features, the prediction model outputs a high predicted execution score for this candidate object.

[0044] In this embodiment, the determination of the feature object is completed by the prediction model. Through the prediction model, the predicted execution scores of multiple target tasks can be obtained, and the predicted recommendation scores corresponding to a certain candidate object for multiple target tasks can be integrated. In some embodiments, the predicted recommendation score may be the weighted sum of the predicted execution scores of multiple target tasks. Based on the obtained predicted recommendation scores of the candidate objects, the N candidate objects with the highest predicted recommendation scores can be determined as the recommended objects, so as to recommend to the target user corresponding to at least two business scenarios. Exemplarily, the value of N can be determined in advance or based on the user tags of the target user. For example, when the target user has the first user tag for some business scenarios and the second user tag for some other business scenarios, one candidate object with the highest predicted recommendation score is determined as the recommended object in the business scenario corresponding to the first user tag, and one candidate object with the highest predicted recommendation score is also determined as the recommended object in the business scenario corresponding to the second user tag. When the target user only has the first user tag for some business scenarios and does not have the second user tag, two candidate objects with the highest predicted recommendation scores are determined as the recommended objects in the business scenario corresponding to the first user tag.

[0045] According to some embodiments, the multiple target tasks include multiple tasks that can be executed in sequence, and the subsequent target tasks can be executed in response to the completion of the execution of the previous target tasks.

[0046] In some embodiments, the multiple target tasks may correspond to multiple stages in implementing an entire conversion chain. Exemplarily, in some business scenarios corresponding to object recommendation, the multiple target tasks may include: receiving an object, completing the content indicated by the object and receiving a reward, and placing an order using the reward. Thus, through multiple target tasks, full-chain object recommendation can be achieved.

[0047] Figure 3 FIG. 4 shows a flowchart of an object recommendation method 300 according to an exemplary embodiment of the present disclosure. As Figure 3 shown, the object recommendation method 300 includes the following steps:

[0048] Step S302, determining a target user;

[0049] Step S304, determining whether the target user has historical behavior information; the historical behavior information may be the historical purchase information described above, and may only correspond to a certain time period, for example, whether there is historical purchase information in the most recent year. The historical behavior information may correspond to one or more business scenarios, or may correspond to all business scenarios;

[0050] If it is determined that the target user does not have historical behavior information, then step S306 is executed to directly determine the recommended object for the target user based on a preset rule; the preset rule may be to directly recommend the corresponding new user recruitment object, or to directly recommend a popular object with a relatively high purchase probability for the user;

[0051] If it is determined that the target user has historical behavior information, according to this embodiment, for the first business scenario, step S308 is executed to determine that the user label of the target user for the first business scenario is the first user label (such as a new user recruitment label); for the second business scenario, step S310 is executed to determine that the user label of the target user for the second business scenario is the second user label (such as a repeat purchase label); for the third business scenario, step S312 is executed to determine that the user label of the target user for the third business scenario is the second user label;

[0052] Step S314, based on the first user label, determining candidate objects in the first business scenario; the candidate objects are candidate objects corresponding to the first user label in the first business scenario. For example, when the first user label is a new user recruitment label, the candidate objects are new user recruitment candidate objects in the first business scenario;

[0053] Step S316, based on the second user label, determining candidate objects in the second business scenario; the candidate objects are candidate objects corresponding to the second user label in the second business scenario. For example, when the second user label is a repeat purchase label, the candidate objects are repeat purchase candidate objects in the second business scenario;

[0054] Step S318: Determine candidate objects in the third business scenario based on the second user tag; the candidate objects are candidate objects corresponding to the second user tag in the third business scenario. For example, when the second user tag is a repeat purchase tag, the candidate objects are repeat purchase candidate objects in the third business scenario.

[0055] Step S320: Determine the recommendation features of the target user for the first business scenario; in some embodiments, the user features of the target user, the business scenario features of the first business scenario, and the object features of the candidate objects in the first business scenario can be concatenated to determine the recommendation features of the target user for the first business scenario.

[0056] Step S322: Determine the recommendation features of the target user for the second and third business scenarios; in some embodiments, the user features of the target user, the business scenario features of the second and third business scenarios, and the object features of the candidate objects in the second and third business scenarios can be concatenated to determine the recommendation features of the target user for the second and third business scenarios.

[0057] Step S324: Determine the first recommended object of the target user for the first business scenario; in this embodiment, the first recommended object is determined by inputting the recommendation features of the target user for the first business scenario into a recommendation model, and the recommendation model can be a model trained with sample data.

[0058] Step S326: Determine the second recommended object of the target user for the second and third business scenarios; in this embodiment, the second recommended object is determined by inputting the recommendation features of the target user for the second and third business scenarios into a recommendation model, and the recommendation model can be a model trained with sample data. The prediction model can be different from or the same as the prediction model used in step S324. The prediction model can be trained with sample data of one, multiple, or all business scenarios.

[0059] In this embodiment, the first recommended object is at least one of the candidate objects in the first business scenario, and the second recommended object is at least one of the candidate objects in the second and third business scenarios. The first recommended object and the second recommended object are recommended to the target user simultaneously.

[0060] In some embodiments, there may be more business scenarios in addition to the first, second, and third business scenarios, and more user tags in addition to the first and second user tags. The user tags can also be combined with the business scenarios in other ways. For example, if the user tags of the target user for the first, second, and third business scenarios are all the first user tag, then candidate objects in the first, second, and third business scenarios are determined respectively through similar steps. After that, the recommended features for the target user for the first, second, and third business scenarios are determined, and the recommended features are input into the recommendation model to obtain the recommended objects for the target user for the first, second, and third business scenarios.

[0061] According to some embodiments, the recommended object can be output by the recommendation model. Combining Figure 1 with step S104, based on the recommended features, determining the recommended objects for the target user for multiple business scenarios includes: inputting the recommended features into the recommendation model and obtaining the recommended objects for multiple users for multiple business scenarios output by the recommendation model.

[0062] Figure 4 FIG. shows a flowchart of a method 400 for training a recommendation model according to an exemplary embodiment of the present disclosure. As Figure 4 shown, the method 400 for training the recommendation model includes: step S402, obtaining sample user tags of a sample user for multiple sample business scenarios, and the sample user tags of the sample user for multiple sample business scenarios are the same; step S404, obtaining sample objects of each sample business scenario and the true recommendation scores of each sample object for the sample user; step S406, determining sample recommendation features of the sample user for multiple sample business scenarios based on the user features of the sample user and the object features of the sample objects of each sample business scenario; step S408, inputting the sample recommendation features into the recommendation model and obtaining the predicted recommendation scores of each sample object for the sample user output by the recommendation model; step S410, calculating a loss value based on the respective true recommendation scores and predicted recommendation scores of multiple sample objects of multiple sample business scenarios; and step S412, adjusting the parameters of the recommendation model based on the loss value.

[0063] In this embodiment, the method 400 for training the recommendation model can be performed iteratively multiple times until the loss value is less than a pre-determined threshold. The adjusted recommendation model can be used as the recommendation model in any of the above methods to obtain the recommended objects by inputting the recommended features into the recommendation model.

[0064] According to some embodiments, the true recommendation score for each sample object in the method 400 for training a recommendation model includes the true execution scores of multiple sample tasks corresponding to each sample object, where the true execution score is used to characterize the probability that the corresponding sample task is executed by the sample user, and wherein the recommendation model can output other intermediate prediction scores, and the final predicted recommendation score can be obtained based on the intermediate prediction scores. For example, the output of the recommendation model includes the predicted execution scores of multiple sample tasks corresponding to each sample object, and the predicted recommendation score is obtained based on the predicted execution scores of multiple sample tasks corresponding to each sample object. In some embodiments, the predicted recommendation score can be a weighted sum of the predicted execution scores.

[0065] In this embodiment, the method 400 for training a recommendation model is executed for multiple sample tasks, and the multiple sample tasks can be multiple tasks that can be executed in sequence, and the subsequent target task can be executed in response to the completion of the execution of the previous sample task. In some embodiments, the multiple sample tasks can correspond to multiple stages in a whole conversion link. Exemplarily, in some business scenarios corresponding to object recommendation, the multiple sample tasks can include: receiving an object, completing the content indicated by the object and receiving a reward, and placing an order using the reward.

[0066] According to another aspect of the present disclosure, there is also provided an object recommendation device. Figure 5 The structural block diagram of the object recommendation device 500 according to an exemplary embodiment of the present disclosure is shown. As Figure 5 shown, the object recommendation device 500 includes: a feature splicing unit 501 configured to, in response to determining that the user tags of the target user for multiple business scenarios are the same, determine the recommendation features of the target user for multiple business scenarios based on the user features of the target user and the object features of the candidate objects for each business scenario; and an object recommendation unit 502 configured to determine the recommended objects of the target user for multiple business scenarios based on the recommendation features, where the recommended objects are determined from the candidate objects for multiple business scenarios.

[0067] Thus, it is possible to perform screening according to different business scenarios, specifically splice the user features of the target user and the candidate object features in different business scenarios, and improve the pertinence and accuracy of object recommendation.

[0068] According to some embodiments, the object recommendation device 500 further includes: a candidate object selection unit configured to determine the candidate objects for each business scenario based on the user tags of the target user for each business scenario. The specific determination method can be the same as that described above and will not be elaborated here.

[0069] According to some embodiments, the candidate object selection unit is further configured to determine, based on user tags and the historical behavior information of the target user in this business scenario, candidate objects of the target user for this business scenario, where the candidate objects of the target user for this business scenario are objects in this business scenario that have not been claimed by the target user.

[0070] Exemplarily, the historical behavior information may include whether the target user has claimed this object in the past. If the target user has claimed this object in the past, this object is removed from at least one candidate object, so that at least one candidate object of the target user for this business scenario is an object in this business scenario that has not been claimed by the target user. This ensures that the target user will not be recommended the same object multiple times.

[0071] According to some embodiments, the user tags of the target user for each business scenario are determined based on the historical behavior information of the target user in this business scenario. The determined user tags may represent the categories of the historical behavior information of the target user in this business scenario.

[0072] Further, the historical behavior information may include historical purchase information. The object recommendation apparatus 500 further includes: a user tag generation unit configured to: in response to determining that the target user has no historical purchase information in this business scenario, determine the user tag as a new customer tag; and in response to determining that the target user has historical purchase information in this business scenario, determine the user tag as a repeat purchase tag. In some embodiments, according to the frequency of the purchase behavior of the target user, the repeat purchase tag may also be determined as a high-frequency repeat purchase tag and a low-frequency repeat purchase tag. The specific determination method may be the same as that described above and will not be elaborated here.

[0073] Thus, the target user can be screened for different business scenarios for targeted recommendation.

[0074] According to some embodiments, the object recommendation apparatus 500 further includes: a candidate object screening unit configured to, before determining the user tags of the target user for each business scenario, in response to determining that the target user has no historical behavior information in each business scenario within a predetermined time range, determine the recommended objects of the target user based on a preset rule. Specifically, the historical behavior information may be historical purchase information. The preset rule may be to directly recommend the object corresponding to the new customer tag, or directly recommend the popular objects with a relatively high purchase probability of the user.

[0075] According to some embodiments, the object recommendation unit 502 in the object recommendation device 500 is further configured to: based on the recommendation features, determine the predicted execution scores of multiple target tasks corresponding to the candidate objects, where the predicted execution scores are used to characterize the probability that the corresponding target tasks are executed by the target user; based on the predicted execution scores of the multiple target tasks corresponding to each candidate object, determine the predicted recommendation scores corresponding to each candidate object; and based on the predicted recommendation scores corresponding to at least one candidate object, determine the recommended objects for the target user for multiple service scenarios from the candidate objects.

[0076] In this embodiment, the predicted execution score is used to characterize the probability that the corresponding target task is executed by the target user. The determination of the feature object is completed through a prediction model. Through the prediction model, the predicted execution scores of multiple target tasks can be obtained, and the predicted recommendation scores corresponding to a certain candidate object for multiple target tasks can be integrated. The specific determination method can be the same as that described above and will not be elaborated here.

[0077] According to some embodiments, the multiple target tasks include multiple tasks that can be executed in sequence, and the subsequent target task can be executed in response to the completion of the execution of the preceding target task. As described above, in some embodiments, the multiple target tasks can correspond to multiple stages in a whole conversion link. Thus, through the multiple target tasks, object recommendation for the whole link can be achieved.

[0078] According to some embodiments, the object recommendation unit is further configured to: input the recommendation features into a recommendation model and obtain the recommended objects for the target user for multiple service scenarios output by the recommendation model.

[0079] According to another aspect of the present disclosure, a training device 600 for a recommendation model is further provided. Figure 6 The structural block diagram of the training device 600 for the recommendation model according to an exemplary embodiment of the present disclosure is shown. As Figure 6As shown, the apparatus 600 for training a recommendation model includes: a data acquisition unit 601 configured to acquire sample user tags of a sample user for a plurality of sample service scenarios, and the sample user tags of the sample user for the plurality of sample service scenarios are the same; a score acquisition unit 602 configured to acquire sample objects of each sample service scenario and true recommendation scores of each sample object for the sample user; a feature determination unit 603 configured to determine sample recommendation features of the sample user for the plurality of sample service scenarios based on user features of the sample user and object features of sample objects of each sample service scenario; a prediction scoring unit 604 configured to input the sample recommendation features into the recommendation model and acquire predicted recommendation scores of each sample object for the sample user output by the recommendation model; a loss calculation unit 605 configured to calculate a loss value based on the true recommendation scores and predicted recommendation scores of the plurality of sample objects corresponding to the plurality of sample service scenarios; and a model adjustment unit 606 configured to adjust parameters of the recommendation model based on the loss value.

[0080] According to some embodiments, the true recommendation score of each sample object for the sample user includes true execution scores of a plurality of sample tasks corresponding to each sample object, the true execution scores being used to represent the probability that the corresponding sample task is executed by the sample user, and the output of the recommendation model includes predicted execution scores of a plurality of sample tasks corresponding to each sample object, and the predicted recommendation score is obtained based on the predicted execution scores of the plurality of sample tasks corresponding to each sample object.

[0081] In this embodiment, the apparatus 600 for training the recommendation model can be executed multiple times in an iterative manner until the loss value is less than a pre-determined threshold. The apparatus 600 for training the recommendation model executes for a plurality of sample tasks, and the plurality of sample tasks can be a plurality of tasks that can be executed in sequence, and the subsequent target task can be executed in response to the completion of the execution of the previous sample task.

[0082] According to another aspect of the present disclosure, there is also provided an electronic device, including: at least one processor; and a memory, where the memory stores a program, and the program includes instructions that, when executed by the processor, enable the processor to execute the above object recommendation method.

[0083] According to another aspect of the present disclosure, there is also provided a non-transitory computer-readable storage medium storing a computer program, where the computer program includes instructions that, when executed by a processor of a computing device, cause the computer to execute the above object recommendation method.

[0084] According to another aspect of the present disclosure, there is also provided a computer program product including a computer program, wherein when the computer program is executed by a processor, the steps of the above-described object recommendation method are implemented.

[0085] Refer to Figure 7 , the block diagram of an electronic device 700 that can be used as the present disclosure will now be described. It is an example of a hardware device that can be applied to various aspects of the present disclosure. The electronic device can be different types of computer devices, such as a laptop computer, a desktop computer, a workbench, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as a personal digital processor, a cellular phone, a smart phone, a wearable device, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are only examples and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0086] Figure 7 The block diagram of an electronic device according to an embodiment of the present disclosure is shown. As Figure 7 shown, the electronic device 700 may include at least one processor 701, a working memory 702, an I / O device 704, a display device 705, a storage device 706, and a communication interface 707 that can communicate with each other through a system bus 703.

[0087] The processor 701 can be a single processing unit or multiple processing units, and all processing units can include a single or multiple computing units or multiple cores. The processor 701 can be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, state machines, logic circuits, and / or any device that manipulates signals based on operation instructions. The processor 701 can be configured to obtain and execute computer-readable instructions stored in the working memory 702, the storage device 706, or other computer-readable media, such as the program code of an operating system 702a, the program code of an application 702b, and the like.

[0088] The working memory 702 and the storage device 706 are examples of computer-readable storage media for storing instructions that are executed by the processor 701 to implement the various functions described above. The working memory 702 may include both volatile and non-volatile memory (e.g., RAM, ROM, etc.). In addition, the storage device 706 may include a hard disk drive, a solid state drive, removable media, including external and removable drives, memory cards, flash memory, floppy disks, optical discs (e.g., CD, DVD), storage arrays, network-attached storage, storage area networks, and the like. The working memory 702 and the storage device 706 may both be collectively referred to herein as memory or computer-readable storage media and may be non-transitory media capable of storing computer-readable, processor-executable program instructions as computer program code that can be executed by the processor 701 as a particular machine configured to implement the operations and functions described in the examples herein.

[0089] The I / O device 704 may include input devices and / or output devices. The input devices may be any type of device capable of inputting information to the electronic device 700 and may include, but are not limited to, a mouse, a keyboard, a touch screen, a trackpad, a trackball, a joystick, a microphone, and / or a remote control. The output devices may be any type of device capable of presenting information and may include, but are not limited to, a video / audio output terminal, a vibrator, and / or a printer.

[0090] The communication interface 707 allows the electronic device 700 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunications networks and may include, but is not limited to, a modem, a network card, an infrared communication device, a wireless communication transceiver, and / or a chipset, such as a BluetoothTM device, an 802.11 device, a WiFi device, a WiMax device, a cellular communication device, and / or the like.

[0091] The application program 702b in the working register 702 may be loaded and executed to perform the various methods and processes described above, such as Figure 1 steps S101 - S104 in. In some embodiments, part or all of the computer program may be loaded and / or installed onto the electronic device 700 via the storage device 706 and / or the communication interface 707. When the computer program is loaded and executed by the processor 701, one or more steps of the data processing method described above may be performed.

[0092] The various embodiments of the systems and techniques described above in this specification can be implemented in digital electronic circuitry, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on a chip (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be a special-purpose or general-purpose programmable processor that receives data and instructions from, and transmits data and instructions to, a storage system, at least one input device, and at least one output device.

[0093] The program code for implementing the methods of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the program codes, when executed by the processor or controller, cause the functions / operations specified in the flowchart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partly on the machine, as a stand-alone software package partly on the machine and partly on a remote machine, or entirely on the remote machine or server.

[0094] In the context of the present disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0095] To provide for interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can also be used to provide for interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0096] The systems and techniques described herein can be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local area network (LAN), wide area network (WAN), and the Internet.

[0097] A computing system can include a client and a server. The client and the server are generally remote from each other and typically interact through a communication network. The client - server relationship is created by computer programs that run on the respective computers and have a client - server relationship with each other.

[0098] It should be understood that the various forms of the processes shown above can be reordered, added to, or deleted. For example, the steps recited in this disclosure can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and no limitation is made herein.

[0099] Although embodiments or examples of the present disclosure have been described with reference to the accompanying drawings, it should be understood that the above methods, systems, and devices are merely exemplary embodiments or examples, and the scope of the present invention is not limited by these embodiments or examples, but is only defined by the authorized claims and their equivalent scope. Various elements in the embodiments or examples may be omitted or replaced by their equivalent elements. In addition, the steps may be executed in an order different from that described in the present disclosure. Further, the various elements in the embodiments or examples may be combined in various ways. Importantly, with the evolution of technology, many of the elements described herein may be replaced by equivalent elements that emerge after the present disclosure.

Claims

1. A method for object recommendation, comprising: In response to determining that the user tags of a target user for multiple business scenarios including a first business scenario and a second business scenario are the same, based on the user characteristics of the target user and the object characteristics of candidate objects for each business scenario, determining the recommendation characteristics of the target user for the multiple business scenarios, including: Determining a first business scenario feature of the target user for the first business scenario and a first object feature of a candidate object in the first business scenario, and a second business scenario feature of the target user for the second business scenario and a second object feature of a candidate object in the second business scenario; and Concatenating the user characteristics of the target user, the first business scenario feature, the second business scenario feature, the first object feature, and the second object feature to determine the recommendation characteristics; and Based on the recommendation characteristics, determining recommended objects of the target user for the multiple business scenarios, where the recommended objects are determined from candidate objects of the multiple business scenarios.

2. The method according to claim 1, wherein, Based on the recommendation characteristics, determining recommended objects of the target user for the multiple business scenarios, including: Based on the recommendation characteristics, obtaining predicted execution scores of multiple target tasks corresponding to the candidate objects, where the predicted execution scores are used to represent the probability that the corresponding target tasks are executed by the target user; Based on the predicted execution scores of the multiple target tasks corresponding to each candidate object, determining predicted recommendation scores corresponding to each candidate object; and Based on the predicted recommendation scores corresponding to each candidate object, determining the recommended objects of the target user for the multiple business scenarios from the candidate objects.

3. The method according to claim 2, wherein, The multiple target tasks include multiple tasks that can be executed in sequence, and subsequent target tasks can be executed in response to the completion of the execution of previous target tasks.

4. The method according to claim 1, further comprising: Based on the user tags of the target user for each business scenario, determining candidate objects for each business scenario.

5. The method according to claim 4, wherein Based on the user tags of the target user for each business scenario, determining candidate objects for each business scenario, further comprising: Based on the user tags and the historical behavior information of the target user in this business scenario, determining candidate objects of the target user for this business scenario, where the candidate objects of the target user for this business scenario are objects in this business scenario that have not been claimed by the target user.

6. The method according to claim 1, wherein, The user tags of the target user for each business scenario are determined based on the historical behavior information of the target user in this business scenario.

7. The method according to claim 6, the method further comprising: In response to determining that the target user does not have historical behavior information in each business scenario within a predetermined time range, determining the recommended objects of the target user based on a preset rule.

8. The method according to claim 6, wherein The historical behavior information includes historical purchase information, where determining the user tags of the target user for each business scenario includes: In response to determining that the target user has no historical purchase information in this business scenario, determine that the user label is a new customer acquisition label; and In response to determining that the target user has historical purchase information in this business scenario, determine that the user label is a repeat purchase label.

9. The method according to claim 1, wherein Based on the recommendation features, determine the recommended objects for the target user for the multiple business scenarios, including: Input the recommendation features into a recommendation model, and obtain the recommended objects for the target user for the multiple business scenarios output by the recommendation model.

10. The method according to claim 9, wherein The training method of the recommendation model includes: Obtain the sample user labels of a sample user for multiple sample business scenarios, and the sample user labels of the sample user for the multiple sample business scenarios are the same; Obtain the sample objects of each sample business scenario and the true recommendation scores of each sample object for the sample user; Based on the user features of the sample user and the object features of the sample objects of each sample business scenario, determine the sample recommendation features of the sample user for the multiple sample business scenarios; Input the sample recommendation features into the recommendation model and obtain the predicted recommendation scores of each sample object for the sample user output by the recommendation model; Based on the true recommendation scores and predicted recommendation scores of the multiple sample objects of the multiple sample business scenarios respectively, calculate a loss value; and Adjust the parameters of the recommendation model based on the loss value.

11. The method according to claim 10, wherein, The true recommendation score of each sample object for the sample user includes the true execution scores of multiple sample tasks corresponding to each sample object, and the true execution score is used to represent the probability that the corresponding sample task is executed by the sample user, and wherein, the output of the recommendation model includes the predicted execution scores of multiple sample tasks corresponding to each sample object, wherein, the predicted recommendation score is obtained based on the predicted execution scores of multiple sample tasks corresponding to each sample object.

12. An apparatus for object recommendation, including: A feature splicing unit, configured to, in response to determining that the user labels of a target user for multiple business scenarios including a first business scenario and a second business scenario are the same, based on the user features of the target user and the object features of the candidate objects of each business scenario, determine the recommendation features of the target user for the multiple business scenarios, including: Determine the first business scenario feature of the target user for the first business scenario and the first object feature of the candidate object in the first business scenario, and the second business scenario feature of the target user for the second business scenario and the second object feature of the candidate object in the second business scenario; and Splice the user features, the first business scenario feature, the second business scenario feature, the first object feature, and the second object feature of the target user to determine the recommendation features; and An object recommendation unit, configured to determine the recommended objects for the target user for the multiple business scenarios based on the recommendation features, and the recommended objects are determined from the candidate objects of the multiple business scenarios.

13. The apparatus according to claim 12, wherein, The object recommendation unit is further configured to: Based on the recommendation features, determine the predicted execution scores of multiple target tasks corresponding to the candidate objects, where the predicted execution scores are used to characterize the probabilities that the corresponding target tasks are executed by the target user; Based on the predicted execution scores of the multiple target tasks corresponding to each candidate object, determine the predicted recommendation scores corresponding to each candidate object; And Based on the predicted recommendation scores corresponding to each candidate object, determine the recommended objects for the target user for the multiple service scenarios from the candidate objects.

14. The device according to claim 13, wherein, The multiple target tasks include multiple tasks that can be executed in sequence, and the subsequent target tasks can be executed in response to the completion of the execution of the previous target tasks.

15. The apparatus according to claim 12, further comprising: A candidate object selection unit, configured to determine candidate objects for each service scenario based on the user tags of the target user for each service scenario.

16. The device according to claim 15, wherein, The candidate object selection unit is further configured to: Based on the user tags and the historical behavior information of the target user in this service scenario, determine the candidate objects for the target user in this service scenario, where the candidate objects for the target user in this service scenario are the objects that have not been received by the target user in this service scenario.

17. The device according to claim 12, wherein The user tags of the target user for each service scenario are determined based on the historical behavior information of the target user in this service scenario.

18. The apparatus according to claim 17, further comprising: A candidate object screening unit, configured to, before determining the user tags of the target user for each service scenario, in response to determining that the target user does not have historical behavior information in each service scenario within a predetermined time range, determine the recommended objects of the target user based on preset rules.

19. The device according to claim 17, wherein, The historical behavior information includes historical purchase information, and the apparatus further comprises: A user tag generation unit, configured to: In response to determining that the target user does not have historical purchase information in this service scenario, determine that the user tag is a new customer acquisition tag; and In response to determining that the target user has historical purchase information in this service scenario, determine that the user tag is a repeat purchase tag.

20. The apparatus according to claim 12, wherein, The object recommendation unit is further configured to: Input the recommendation features into a recommendation model, and obtain the recommended objects for the target user for the multiple service scenarios output by the recommendation model.

21. The apparatus according to claim 20, wherein The training apparatus of the recommendation model includes: A data acquisition unit, configured to acquire sample user tags of a sample user for multiple sample service scenarios, and the sample user tags of the sample user for the multiple sample service scenarios are the same; A score acquisition unit, configured to acquire sample objects for each sample service scenario and the true recommendation scores of each sample object for the sample user; A feature determination unit, configured to determine sample recommendation features of the sample user for the multiple sample service scenarios based on the user features of the sample user and the object features of the sample objects for each sample service scenario. A prediction scoring unit, configured to input the sample recommendation features into a recommendation model and obtain the predicted recommendation scores of each sample object for the sample user output by the recommendation model; A loss calculation unit, configured to calculate a loss value based on the true recommendation scores and the predicted recommendation scores of the respective sample objects of the multiple sample business scenarios; and A model adjustment unit, configured to adjust the parameters of the recommendation model based on the loss value.

22. The apparatus according to claim 21, wherein, The true recommendation score of each sample object for the sample user includes the true execution scores of multiple sample tasks corresponding to each sample object, and the true execution score is used to characterize the probability that the corresponding sample task is executed by the sample user. And wherein, the output of the recommendation model includes the predicted execution scores of multiple sample tasks corresponding to each sample object. Wherein, the predicted recommendation score is obtained based on the predicted execution scores of multiple sample tasks corresponding to each sample object.

23. An electronic device, comprising: A processor; And A memory, the memory storing a program, the program including instructions that, when executed by the processor, cause the processor to execute the method according to any one of claims 1-11.

24. A non-transitory computer-readable storage medium, the non-transitory computer-readable storage medium storing a computer program, the computer program including instructions that, when executed by a processor of a computing device, cause the computing device to execute the method according to any one of claims 1-11.

25. A computer program product comprising a computer program, wherein, The steps of the method according to any one of claims 1-11 are implemented when the computer program is executed by a processor.

Images