Recommendation method, device, computer readable storage medium and electronic device

Abstract

The embodiments of the present disclosure disclose a recommendation method, device, computer-readable storage medium and electronic device. The method includes: using multiple recall strategies to recall materials for a first user to obtain multiple recall results; judging whether there is recommended material behavior performance information of the first user in the historical behavior data to obtain a judgment result; according to the judgment result, determining multiple weights of multiple recall strategies for the first user in a corresponding weight determination manner; using multiple weights to fuse multiple recall results to obtain a fusion result; and recommending materials to the first user according to the fusion result. The embodiments of the present disclosure can improve the personalization of the recommendation results when recommending materials, so as to achieve accurate personalized recommendations for users.

Description

Technical Field

[0001] The present disclosure relates to the technical field of information recommendation, and in particular to a recommendation method, device, computer-readable storage medium and electronic device. Background Art

[0002] At present, the use of recommendation systems is very common. Generally speaking, when recommending materials, it is necessary to first use multiple recall strategies to recall materials separately to obtain multiple recall results, and then make material recommendations based on the obtained recall results. However, when using existing recommendation methods, the personalization of recommendation results is low, and it is impossible to achieve accurate personalized recommendations for users. Summary of the invention

[0003] In order to solve the above technical problems, the present disclosure is proposed. The embodiments of the present disclosure provide a recommendation method, device, computer-readable storage medium and electronic device.

[0004] According to one aspect of an embodiment of the present disclosure, a recommendation method is provided, including:

[0005] Use multiple recall strategies to recall materials for the first user to obtain multiple recall results;

[0006] Determine whether there is recommended material behavior performance information of the first user in the historical behavior data to obtain a determination result;

[0007] According to the judgment result, determining a plurality of weights corresponding to the plurality of recall strategies for the first user in a corresponding weight determination manner;

[0008] Using the multiple weights, fusing the multiple recall results to obtain a fusion result;

[0009] According to the fusion result, material is recommended to the first user.

[0010] In an optional example, determining, according to the judgment result, a plurality of weights corresponding to the plurality of recall strategies for the first user in a corresponding weight determination manner includes:

[0011] In a case where the judgment result indicates that there is recommended material behavior performance information of the first user in the historical behavior data, determining a plurality of weights corresponding to the plurality of recall strategies for the first user according to the recommended material behavior performance information of the first user;

[0012] When the judgment result indicates that there is no recommended material behavior performance information of the first user in the historical behavior data, multiple weights corresponding to the multiple recall strategies are determined for the first user according to a preset corresponding relationship; wherein the preset corresponding relationship is a corresponding relationship between the recall strategy and the weight.

[0013] In an optional example, determining multiple weights corresponding to the multiple recall strategies for the first user according to the recommended material behavior performance information of the first user includes:

[0014] Determine a first type of recall strategy among the multiple recall strategies; wherein the recommended material behavior performance information of the first user includes recommended material behavior performance information related to each recall strategy in the first type of recall strategy;

[0015] Determine the number of material exposures and the number of material views corresponding to each recall strategy in the first category of recall strategies according to the recommended material behavior performance information related to each recall strategy in the first category of recall strategies;

[0016] The weight corresponding to each recall strategy in the first category of recall strategies is determined according to the number of material exposures and the number of material views corresponding to each recall strategy in the first category of recall strategies.

[0017] In an optional example, determining the weight corresponding to each recall strategy in the first category of recall strategies according to the number of material exposures and the number of material views corresponding to each recall strategy in the first category of recall strategies includes:

[0018] Calculating the sum of the material exposure times corresponding to various recall strategies in the first type of recall strategies;

[0019] Calculating a ratio of the sum to the number of material exposures corresponding to any recall strategy in the first category of recall strategies;

[0020] Using the formula Calculate the weight corresponding to any of the recall strategies; wherein Q is the weight corresponding to any of the recall strategies, V is the number of material views corresponding to any of the recall strategies, and S is a preset value. is the ratio.

[0021] In an optional example, determining multiple weights corresponding to the multiple recall strategies for the first user according to the recommended material behavior performance information of the first user includes:

[0022] Determine a first type of recall strategy and a second type of recall strategy among the multiple recall strategies; wherein the recommended material behavior performance information of the first user includes the recommended material behavior performance information related to each recall strategy in the first type of recall strategy, and the recommended material behavior performance information of the first user does not include the recommended material behavior performance information related to each recall strategy in the second type of recall strategy;

[0023] Determining a weight corresponding to each recall strategy in the first category of recall strategies according to the recommended material behavior performance information related to each recall strategy in the first category of recall strategies;

[0024] According to the weight corresponding to each recall strategy in the first category of recall strategies, the weight corresponding to each recall strategy in the second category of recall strategies is determined.

[0025] In an optional example, determining the weight corresponding to each recall strategy in the second category of recall strategies according to the weight corresponding to each recall strategy in the first category of recall strategies includes:

[0026] Selecting the weight with the largest value from the weights corresponding to the various recall strategies in the first category of recall strategies;

[0027] Determine the weight corresponding to each recall strategy in the second type of recall strategies; wherein each determined weight is greater than the weight with the largest value.

[0028] In an alternative example,

[0029] The step of determining, for the first user, a plurality of weights corresponding to the plurality of recall strategies based on the recommended material behavior performance information of the first user includes:

[0030] Determining a current recommended business scenario for the first user;

[0031] Determining, for the first user, a plurality of weights corresponding to the plurality of recall strategies according to the recommended material behavior performance information of the first user related to the current recommended business scenario;

[0032] and / or,

[0033] The step of determining, for the first user, a plurality of weights corresponding to the plurality of recall strategies according to the preset corresponding relationship includes:

[0034] Determining a current recommended business scenario for the first user;

[0035] According to the preset corresponding relationship corresponding to the current recommended business scenario, multiple weights corresponding to the multiple recall strategies are determined for the first user.

[0036] In an optional example, after recommending materials to the first user according to the fusion result, the method further includes:

[0037] Determining the first user's behavior performance on the material recommendation result;

[0038] The historical behavior data is updated according to the behavior performance.

[0039] According to another aspect of an embodiment of the present disclosure, there is provided a recommendation device, including:

[0040] A first acquisition module, used to adopt multiple recall strategies to recall materials for the first user to obtain multiple recall results;

[0041] A second acquisition module is used to determine whether there is the recommended material behavior performance information of the first user in the historical behavior data to obtain a determination result;

[0042] A first determination module, configured to determine, according to the judgment result and in a corresponding weight determination manner, a plurality of weights corresponding to the plurality of recall strategies for the first user;

[0043] A third acquisition module, configured to fuse the multiple recall results using the multiple weights to obtain a fusion result;

[0044] A recommendation module is used to recommend materials to the first user based on the fusion result.

[0045] In an optional example, the first determining module includes:

[0046] A first determination submodule is configured to determine, when the judgment result indicates that there is recommended material behavior performance information of the first user in the historical behavior data, a plurality of weights corresponding to the plurality of recall strategies for the first user according to the recommended material behavior performance information of the first user;

[0047] The second determination submodule is used to determine multiple weights corresponding to the multiple recall strategies for the first user according to a preset correspondence relationship when the judgment result indicates that there is no recommended material behavior performance information of the first user in the historical behavior data; wherein the preset correspondence relationship is a correspondence relationship between the recall strategy and the weight.

[0048] In an optional example, the first determining submodule includes:

[0049] A first determining unit is used to determine a first type of recall strategy among the multiple recall strategies; wherein the recommended material behavior performance information of the first user includes recommended material behavior performance information related to each recall strategy in the first type of recall strategy;

[0050] A second determination unit is used to determine the number of material exposures and the number of material views corresponding to each recall strategy in the first category of recall strategies according to the recommended material behavior performance information related to each recall strategy in the first category of recall strategies;

[0051] The third determination unit is used to determine the weight corresponding to each recall strategy in the first category of recall strategies according to the number of material exposures and the number of material views corresponding to each recall strategy in the first category of recall strategies.

[0052] In an optional example, the third determining unit includes:

[0053] A first calculation subunit, used to calculate the sum of material exposure times corresponding to various recall strategies in the first type of recall strategy;

[0054] A second calculation subunit is used to calculate the ratio of the sum to the number of material exposures corresponding to any recall strategy in the first type of recall strategies;

[0055] The third calculation subunit is used to use the formula Calculate the weight corresponding to any of the recall strategies; wherein Q is the weight corresponding to any of the recall strategies, V is the number of material views corresponding to any of the recall strategies, and S is a preset value. is the ratio.

[0056] In an optional example, the first determining submodule includes:

[0057] a fourth determination unit, configured to determine a first type of recall strategy and a second type of recall strategy among the multiple recall strategies; wherein the recommended material behavior performance information of the first user includes the recommended material behavior performance information related to each recall strategy in the first type of recall strategy, and the recommended material behavior performance information of the first user does not include the recommended material behavior performance information related to each recall strategy in the second type of recall strategy;

[0058] a fifth determining unit, configured to determine a weight corresponding to each recall strategy in the first category of recall strategies according to the recommended material behavior performance information related to each recall strategy in the first category of recall strategies;

[0059] The sixth determining unit is used to determine the weight corresponding to each recall strategy in the second category of recall strategies according to the weight corresponding to each recall strategy in the first category of recall strategies.

[0060] In an optional example, the sixth determining unit includes:

[0061] A screening subunit, configured to screen the weight with the largest value from the weights corresponding to the various recall strategies in the first category of recall strategies;

[0062] A determination subunit is used to determine the weight corresponding to each recall strategy in the second type of recall strategy; wherein each determined weight is greater than the weight with the largest value.

[0063] In an alternative example,

[0064] The first determining submodule includes:

[0065] A seventh determining unit, configured to determine a current recommended service scenario for the first user;

[0066] An eighth determining unit, configured to determine, for the first user, a plurality of weights corresponding to the plurality of recall strategies according to the recommended material behavior performance information of the first user related to the current recommended business scenario;

[0067] If / or,

[0068] The second determining submodule includes:

[0069] A ninth determining unit, configured to determine a current recommended service scenario for the first user;

[0070] The tenth determining unit is used to determine a plurality of weights corresponding to the plurality of recall strategies for the first user according to a preset corresponding relationship corresponding to the current recommended business scenario.

[0071] In an optional example, the device further includes:

[0072] A second determination module is used to determine the behavior performance of the first user on the material recommendation result after recommending the material to the first user according to the fusion result;

[0073] An updating module is used to update the historical behavior data according to the behavior performance.

[0074] According to another aspect of the embodiments of the present disclosure, a computer-readable storage medium is provided, wherein the storage medium stores a computer program, and the computer program is used to execute the above-mentioned recommendation method.

[0075] According to another aspect of the embodiments of the present disclosure, there is provided an electronic device, including:

[0076] processor;

[0077] a memory for storing instructions executable by the processor;

[0078] The processor is used to read the executable instructions from the memory and execute the instructions to implement the above-mentioned recommended method.

[0079] In an embodiment of the present disclosure, when it is necessary to recommend materials to a first user, multiple recall strategies can be firstly adopted to recall materials for the first user to obtain multiple recall results; next, multiple weights corresponding to the multiple recall strategies can be determined for the first user in a corresponding weight determination manner based on a judgment result representing whether there is recommended material behavior performance information of the first user in the historical behavior data; thereafter, the multiple recall results can be fused using the multiple weights determined for the first user to obtain a fused result; thereafter, based on the fused result, material recommendations to the first user can be achieved. It should be noted that when the recommendation method in the relevant technology is adopted, the weight corresponding to each recall strategy is a fixed weight determined in advance. For any recall strategy, the weight adopted by all users is unified. In comparison, in the embodiment of the present disclosure, a suitable weight determination method can be adopted for each user to determine the weight corresponding to each recall strategy. In this way, for any recall strategy, the weight adopted by all users is not unified (for example, for the recall strategy of the business rule dimension, the weight adopted by the first user may be 0.8, while the weight adopted by the second user may be 0.6). That is, the embodiment of the present disclosure can realize the diversification of the weights used in the fusion stage to adapt to different users. Therefore, the embodiment of the present disclosure can improve the personalization of the recommendation results when making material recommendations, so as to achieve accurate personalized recommendations for users.

[0080] The technical solution of the present disclosure is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] The above and other purposes, features and advantages of the present disclosure will become more apparent by describing the embodiments of the present disclosure in more detail in conjunction with the accompanying drawings. The accompanying drawings are used to provide a further understanding of the embodiments of the present disclosure and constitute a part of the specification. Together with the embodiments of the present disclosure, they are used to explain the present disclosure and do not constitute a limitation of the present disclosure. In the drawings, the same reference numerals generally represent the same components or steps.

[0082] Figure 1 It is a flowchart of a recommendation method provided by an exemplary embodiment of the present disclosure.

[0083] Figure 2 It is a flowchart of a recommendation method provided by another exemplary embodiment of the present disclosure.

[0084] Figure 3 It is a flowchart of a recommendation method provided by yet another exemplary embodiment of the present disclosure.

[0085] Figure 4 It is a structural diagram of a recommendation device provided by an exemplary embodiment of the present disclosure.

[0086] Figure 5 It is a structural schematic diagram of a recommendation device provided by another exemplary embodiment of the present disclosure.

[0087] Figure 6 is a structural diagram of an electronic device provided by an exemplary embodiment of the present disclosure. DETAILED DESCRIPTION

[0088] Below, the exemplary embodiments according to the present disclosure will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all the embodiments of the present disclosure, and it should be understood that the present disclosure is not limited to the exemplary embodiments described here.

[0089] It should be noted that the relative arrangement of components and steps, the numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present disclosure unless specifically stated otherwise.

[0090] Those skilled in the art can understand that the terms "first" and "second" in the embodiments of the present disclosure are only used to distinguish different steps, devices or modules, etc., and neither represent any specific technical meaning nor indicate the necessary logical order between them.

[0091] It should also be understood that in the embodiments of the present disclosure, “plurality” may refer to two or more than two, and “at least one” may refer to one, two, or more than two.

[0092] It should also be understood that any component, data or structure mentioned in the embodiments of the present disclosure can generally be understood as one or more, unless explicitly limited or otherwise indicated in the context.

[0093] In addition, the term "and / or" in the present disclosure is only a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in the present disclosure generally indicates that the associated objects before and after are in an "or" relationship.

[0094] It should also be understood that the description of the various embodiments in the present disclosure focuses on the differences between the various embodiments, and the same or similar aspects thereof can be referenced to each other, and for the sake of brevity, they will not be described one by one.

[0095] At the same time, it should be understood that for the convenience of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship.

[0096] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the present disclosure, its application, or uses.

[0097] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered as part of the specification.

[0098] It should be noted that like reference numerals and letters refer to similar items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0099] The disclosed embodiments can be applied to electronic devices such as terminal devices, computer systems, servers, etc., which can operate with many other general or special computing system environments or configurations. Examples of well-known terminal devices, computing systems, environments and / or configurations suitable for use with electronic devices such as terminal devices, computer systems, servers, etc. include, but are not limited to: personal computer systems, server computer systems, thin clients, thick clients, handheld or laptop devices, microprocessor-based systems, set-top boxes, programmable consumer electronics, network personal computers, small computer systems, large computer systems, and distributed cloud computing technology environments including any of the above systems, etc.

[0100] Electronic devices such as terminal devices, computer systems, servers, etc. can be described in the general context of computer system executable instructions (such as program modules) executed by computer systems. Generally, program modules can include routines, programs, object programs, components, logic, data structures, etc., which perform specific tasks or implement specific abstract data types. Computer systems / servers can be implemented in a distributed cloud computing environment, where tasks are performed by remote processing devices linked through a communication network. In a distributed cloud computing environment, program modules can be located on local or remote computing system storage media including storage devices.

[0101] Exemplary Methods

[0102] Figure 1 It is a flowchart of a recommendation method provided by an exemplary embodiment of the present disclosure. Figure 1 The method shown may include step 101, step 102, step 103, step 104 and step 105, and each step is described below.

[0103] Step 101, using multiple recall strategies to recall materials for a first user to obtain multiple recall results.

[0104] Here, the material may be a house source, and the first user may be a user who has a need to buy or rent a house. Of course, the types of the material and the first user are not limited to this. For example, the material may be a book, and the first user may be a user who has a need to buy a book. For ease of understanding, the embodiments of the present disclosure are described by taking the case where the material is a house source and the first user is a user who has a need to buy a house as an example.

[0105] In an embodiment of the present disclosure, the first user can operate on the client corresponding to the service platform provided by the real estate company to initiate a user request (e.g., a new house search request, a second-hand house search request, etc.). The service platform may include a recommendation system. In response to the user request initiated by the first user, the recommendation system may use multiple recall strategies to recall materials for the first user to obtain multiple recall results; wherein, there may be a one-to-one correspondence between the multiple recall results and the multiple recall strategies, and each recall result may include a number (e.g., 10) of housing sources.

[0106] Optionally, the multiple recall strategies include but are not limited to a recall strategy in the dimension of user feature preference, a recall strategy in the dimension of material feature, and a recall strategy in the dimension of business rule.

[0107] The recall strategy of the user feature preference dimension is used to recall materials according to the user preference of the first user. For example, if the first user prefers three-bedroom houses, the houses in the recall results obtained by using the recall strategy may all be three-bedroom houses.

[0108] The recall strategy of the material feature dimension is used to make recommendations based on the contextual browsing of the first user on the service platform. For example, if the first user recently browsed on the service platform a house with a total price of about 6 million and located in Haidian District, the houses in the recall results obtained by using this recall strategy may all be houses with a total price of 5.5 to 6.5 million and located in Haidian District.

[0109] The recall strategy in the business rule dimension is used to recall materials according to pre-defined business rules. For example, the pre-defined business rule may be to recall materials according to the popularity of the listings. Then, the listings in the recall results obtained by using this recall strategy may all be listings with high popularity.

[0110] Step 102, determining whether there is recommended material behavior performance information of the first user in the historical behavior data to obtain a determination result.

[0111] It should be noted that historical behavior data can include a large amount of information. For users who have performed any behavior (such as browsing, searching, sharing, etc.) on the service platform, the historical behavior data can include the behavior data of the user. Specifically, if the user has initiated a user request through the client, and the recommendation system of the service platform recommends a house to the user in response to the user request, the user's behavior data may include the user's recommended material behavior performance information, and the user's recommended material behavior performance information can be used to characterize the user's behavior performance on the houses recommended by the recommendation system, for example, which houses recommended by the recommendation system have not been viewed by the user, which houses recommended by the recommendation system have been viewed by the user and the specific viewing time, etc.

[0112] In step 102, the historical behavior data may be traversed to determine whether there is recommended material behavior performance information of the first user in the historical behavior data, thereby obtaining a determination result.

[0113] Step 103: According to the judgment result, a plurality of weights corresponding to the plurality of recall strategies are determined for the first user in a corresponding weight determination manner.

[0114] Here, there may be a one-to-one correspondence between the multiple recall strategies and the multiple weights.

[0115] It should be noted that there are two possible situations for the judgment result: in one case, the judgment result indicates that there is recommended material behavior performance information of the first user in the historical behavior data. In this case, it can be considered that the first user is a new user for the service platform. In this case, a weight determination method suitable for new users can be used to determine multiple weights of multiple recall strategies for the first user; in another case, the judgment result indicates that there is no recommended material behavior performance information of the first user in the historical behavior data. In this case, it can be considered that the first user is an old user for the service platform. In this case, a weight determination method suitable for old users can be used to determine multiple weights of multiple recall strategies for the first user.

[0116] Step 104, using multiple weights to fuse multiple recall results to obtain a fusion result.

[0117] Here, when multiple weights are used to fuse multiple recall strategies, the fusion strategies used include but are not limited to sequential fusion strategies, weighted fusion strategies, etc. The obtained fusion result may only include some of the properties in the multiple recall results. For example, when the number of recall results is 3 and each recall result includes 10 properties, the fusion result may only include 20 sorted properties. In this way, the process of fusing multiple recall results can also be called a coarse sorting process.

[0118] Step 105: recommend materials to the first user based on the fusion result.

[0119] Here, after obtaining the fusion result including the sorted 20 houses, the fusion result can be further processed using the refined sorting algorithm to obtain the material recommendation result that needs to be finally displayed to the first user. The material recommendation result can only include some houses in the fusion result. For example, the material recommendation result can only include 10 houses arranged in sequence. Afterwards, the material recommendation result can be displayed on the client used by the first user to realize the material recommendation to the first user. Each house in the material recommendation result can be specifically displayed in the form of a house card.

[0120] In an embodiment of the present disclosure, when it is necessary to recommend materials to a first user, multiple recall strategies can be firstly adopted to recall materials for the first user to obtain multiple recall results; next, multiple weights corresponding to the multiple recall strategies can be determined for the first user in a corresponding weight determination manner based on a judgment result representing whether there is recommended material behavior performance information of the first user in the historical behavior data; thereafter, the multiple recall results can be fused using the multiple weights determined for the first user to obtain a fused result; thereafter, based on the fused result, material recommendations to the first user can be achieved. It should be noted that when the recommendation method in the relevant technology is adopted, the weight corresponding to each recall strategy is a fixed weight determined in advance. For any recall strategy, the weight adopted by all users is unified. In comparison, in the embodiment of the present disclosure, a suitable weight determination method can be adopted for each user to determine the weight corresponding to each recall strategy. In this way, for any recall strategy, the weight adopted by all users is not unified (for example, for the recall strategy of the business rule dimension, the weight adopted by the first user may be 0.8, while the weight adopted by the second user may be 0.6). That is, the embodiment of the present disclosure can realize the diversification of the weights used in the fusion stage to adapt to different users. Therefore, the embodiment of the present disclosure can improve the personalization of the recommendation results when making material recommendations, so as to achieve accurate personalized recommendations for users.

[0121] Figure 2 It is a flowchart of a recommendation method provided by another exemplary embodiment of the present disclosure. Figure 2 The method shown may include step 201, step 202, step 203, step 204, step 205 and step 206, and each step is described below.

[0122] Step 201 , using multiple recall strategies to recall materials for a first user to obtain multiple recall results.

[0123] Step 202: determine whether there is recommended material behavior performance information of the first user in the historical behavior data to obtain a determination result.

[0124] It should be noted that the specific implementation process of step 201 to step 202 can refer to the description of step 101 to step 102, which will not be repeated here.

[0125] Step 203, when the judgment result indicates that there is recommended material behavior performance information of the first user in the historical behavior data, multiple weights corresponding to multiple recall strategies are determined for the first user according to the recommended material behavior performance information of the first user.

[0126] In a specific implementation, based on the recommended material behavior performance information of the first user, multiple weights corresponding to multiple recall strategies are determined for the first user, including:

[0127] Determine a first type of recall strategy among multiple recall strategies; wherein the recommended material behavior performance information of the first user includes recommended material behavior performance information related to each recall strategy in the first type of recall strategy;

[0128] According to the recommended material behavior performance information related to each recall strategy in the first category of recall strategies, determine the number of material exposures and the number of material views corresponding to each recall strategy in the first category of recall strategies;

[0129] According to the number of material exposures and the number of material views corresponding to each recall strategy in the first category of recall strategies, a weight corresponding to each recall strategy in the first category of recall strategies is determined.

[0130] Here, the first type of recall strategy among the multiple recall strategies can be determined first. Since the recommended material behavior performance information of the first user includes the recommended material behavior performance information related to each recall strategy in the first type of recall strategy, it can be considered that the recommendation system has previously recommended the house source recalled by each recall strategy in the first type of recall strategy to the first user, and therefore, each recall strategy in the first type of recall strategy can be considered as a used recall strategy.

[0131] Next, the number of material exposures and material views corresponding to each recall strategy in the first category of recall strategies can be determined based on the recommended material behavior performance information related to each recall strategy in the first category of recall strategies; wherein, the number of material exposures corresponding to any recall strategy determined can refer to: the number of times the listings recalled using the recall strategy are recommended to the first user, and each recommendation of one listing to the first user can be counted as one time (each recommendation of one listing to the first user can be considered as one exposure event, and each exposure event has a unique exposure ID); the number of material views corresponding to any recall strategy determined can refer to: the number of times the first user actually performs material viewing operations for the listings recommended to the first user and recalled using the recall strategy. Generally speaking, the first user views the listing by clicking on the listing card, so the number of material views here can also be called the number of material clicks.

[0132] Afterwards, the weight corresponding to each recall strategy in the first category of recall strategies may be determined according to the number of material exposures and the number of material views corresponding to each recall strategy in the first category of recall strategies.

[0133] Optionally, according to the number of material exposures and the number of material views corresponding to each recall strategy in the first category of recall strategies, a weight corresponding to each recall strategy in the first category of recall strategies is determined, including:

[0134] Calculate the total number of material exposures corresponding to various recall strategies in the first category of recall strategies;

[0135] Calculate the ratio of the total to the number of material exposures corresponding to any recall strategy in the first category of recall strategies;

[0136] Using the formula Calculate the weight corresponding to any recall strategy; where Q is the weight corresponding to any recall strategy, V is the number of material views corresponding to any recall strategy, and S is the preset value. Is a ratio.

[0137] Here, the preset value S can be 2, 3, 4 or other values, which are not listed here one by one.

[0138] Assuming that the preset value S is specifically 2, the above formula for calculating the weight of the recall strategy can be transformed into:

[0139]

[0140] Among them, V can be considered as the cumulative return in the bandit algorithm. It can be considered as the estimated return in the bandit algorithm.

[0141] It is understandable that the current bandit algorithm utilizes the idea of ​​a multi-armed bandit (MAB), treating each material as an arm. Each time a material is selected for recall, the reward and benefit of each arm are evaluated as weights. After each recommendation is completed, the weight of each arm is dynamically adjusted based on user behavior feedback. The Bandit algorithm can better balance the exploration and utilization issues.

[0142] Of course, the formula for determining the weight corresponding to each recall strategy in the first category of recall strategies is not limited to this. For example, based on the above formula, coefficients can be added to the two items before and after "+" respectively, and the weight calculation can be performed based on the formula after adding the coefficients. This is also feasible.

[0143] In this implementation, the weight corresponding to each recall strategy in the first category of recall strategies can be conveniently calculated by combining the cumulative return and estimated benefit of each recall strategy in the first category of recall strategies with some simple operations (such as addition, multiplication, division, etc.).

[0144] In another specific implementation, multiple weights corresponding to multiple recall strategies are determined for the first user based on the recommended material behavior performance information of the first user, including:

[0145] Determine a first type of recall strategy and a second type of recall strategy among the multiple recall strategies; wherein the recommended material behavior performance information of the first user includes the recommended material behavior performance information related to each recall strategy in the first type of recall strategy, and the recommended material behavior performance information of the first user does not include the recommended material behavior performance information related to each recall strategy in the second type of recall strategy;

[0146] Determine the weight corresponding to each recall strategy in the first category of recall strategies according to the recommended material behavior performance information related to each recall strategy in the first category of recall strategies;

[0147] According to the weight corresponding to each recall strategy in the first category of recall strategies, the weight corresponding to each recall strategy in the second category of recall strategies is determined.

[0148] Here, the first category and the second category of recall strategies among the multiple recall strategies can be determined first. Since the recommended material behavior performance information of the first user includes the recommended material behavior performance information related to each recall strategy in the first category of recall strategies, and the recommended material behavior performance information of the first user does not include the recommended material behavior performance information related to each recall strategy in the second category of recall strategies, it can be considered that the recommendation system has previously recommended the housing sources recalled by each recall strategy in the first category of recall strategies to the first user, and the recommendation system has not previously recommended the housing sources recalled by each recall strategy in the second category of recall strategies to the first user, then each recall strategy in the first category of recall strategies can be considered as a recall strategy that has been used, and each recall strategy in the second category of recall strategies can be considered as a newly added recall strategy.

[0149] Next, the weight corresponding to each recall strategy in the first category of recall strategies can be determined based on the recommended material behavior performance information related to each recall strategy in the first category of recall strategies. The specific determination method can be referred to the description in the previous implementation manner and will not be repeated here.

[0150] Afterwards, the weight corresponding to each recall strategy in the second category of recall strategies may be determined according to the weight corresponding to each recall strategy in the first category of recall strategies.

[0151] Optionally, determining the weight corresponding to each recall strategy in the second category of recall strategies according to the weight corresponding to each recall strategy in the first category of recall strategies includes:

[0152] From the weights corresponding to various recall strategies in the first category of recall strategies, select the weight with the largest value;

[0153] Determine the weight corresponding to each recall strategy in the second type of recall strategies; wherein each determined weight is greater than the weight with the largest value.

[0154] Assuming that the first category of recall strategies includes two recall strategies, and the weights corresponding to the two recall strategies determined for the first user are Q1 and Q2 respectively, then the weight with the largest value can be screened out from Q1 and Q2. Assuming that the weight with the largest value screened out is Q2, when determining the weight corresponding to each recall strategy in the second category of recall strategies for the first user, each determined weight can be greater than Q2. Through this determination method, the weight corresponding to the newly added recall strategy can be larger than the weight corresponding to the used recall strategy, which can improve the exploratory nature of the recommendation system for the newly added recall strategy, thereby facilitating the recommendation system to better develop the newly added recall strategy.

[0155] In this implementation, the weight corresponding to the newly added recall strategy can be conveniently determined based on the weight corresponding to the used recall strategy.

[0156] In another specific implementation, multiple weights corresponding to multiple recall strategies are determined for the first user based on the recommended material behavior performance information of the first user, including:

[0157] Determine a current recommended business scenario for the first user;

[0158] According to the recommended material behavior performance information of the first user related to the current recommended business scenario, multiple weights corresponding to multiple recall strategies are determined for the first user.

[0159] Here, the recommendation business scenario includes but is not limited to the new house recommendation business scenario, the second-hand house recommendation business scenario, etc. Optionally, the new house recommendation business scenario can be subdivided into the new house homepage recommendation business scenario, the new house details page recommendation business scenario, etc.; the second-hand house recommendation business scenario can be subdivided into the second-hand house homepage recommendation business scenario, the second-hand house details page recommendation business scenario, etc.

[0160] In this implementation, the first user's current recommended business scenario can be determined based on the page the client is on when the first user initiates a user request through the client. For example, if the page the client is on is related to the new house business, then it can be determined that the first user's current recommended business scenario is the new house recommended business scenario.

[0161] After determining the current recommended business scenario of the first user, multiple weights corresponding to multiple recall strategies can be determined for the first user based solely on the first user's recommended material behavior performance information related to the determined recommended business scenario. At this time, the number of material exposures used in the weight calculation above can specifically be the number of material exposures under the determined recommended business scenario, and the number of material views used in the weight calculation above can specifically be the number of material views under the determined recommended business scenario.

[0162] In this implementation, by referring to the current recommended business scenario of the first user to determine the weight for the first user, the determined weight can be adapted to the current recommended business scenario, thereby ensuring the recommendation effect.

[0163] Step 204, when the judgment result indicates that there is no recommended material behavior performance information of the first user in the historical behavior data, multiple weights corresponding to multiple recall strategies are determined for the first user according to a preset corresponding relationship; wherein the preset corresponding relationship is a corresponding relationship between the recall strategy and the weight.

[0164] In a specific implementation, determining multiple weights of multiple recall strategies for the first user according to the preset corresponding relationship may include:

[0165] Determine a current recommended business scenario for the first user;

[0166] According to the preset corresponding relationship corresponding to the current recommendation business scenario, multiple weights corresponding to the multiple recall strategies are determined for the first user.

[0167] Here, the correspondence between multiple recommended business scenarios and corresponding preset correspondences can be pre-stored. After determining the current recommended business scenario of the first user, the preset correspondence corresponding to the current recommended business scenario of the first user can be determined based on the pre-stored correspondence, and then based on the determined preset correspondence, multiple weights corresponding to multiple recall strategies are determined for the first user, so that the determined weights can be adapted to the current recommended business scenario, thereby ensuring the recommendation effect.

[0168] Of course, it is also possible not to pre-store the correspondences between multiple recommended business scenarios and corresponding preset correspondences, but to set only one preset correspondence, and subsequently use only this preset correspondence to determine the weight, which is also feasible.

[0169] Step 205: Use multiple weights to fuse multiple recall results to obtain a fusion result.

[0170] Step 206: Recommend materials to the first user based on the fusion result.

[0171] It should be noted that, steps 205 to 206 may refer to the description of steps 104 to 105 and will not be repeated here.

[0172] In an embodiment of the present disclosure, when it is necessary to recommend materials to a first user, multiple recall strategies can be firstly adopted to recall materials for the first user to obtain multiple recall results; next, based on the judgment result used to characterize whether there is recommended material behavior performance information of the first user in the historical behavior data, multiple weights corresponding to multiple recall strategies can be determined for the first user according to the recommended material behavior performance information of the first user, or multiple weights corresponding to multiple recall strategies can be determined for the first user according to a preset corresponding relationship; thereafter, the multiple weights determined for the first user can be used to fuse the multiple recall results to obtain a fused result; thereafter, based on the fused result, the material recommendation to the first user can be realized. It should be noted that when adopting the recommendation method in the relevant technology, the weight corresponding to each recall strategy is a fixed weight determined in advance. For any recall strategy, the weight adopted by all users is unified. In comparison, in the embodiment of the present disclosure, a suitable weight determination method can be adopted for each user to determine the weight corresponding to each recall strategy. Specifically, for old users, the weight can be determined by using their recommended material behavior performance information, and for new users, the weight can be determined by using a preset corresponding relationship. In this way, for any recall strategy, the weight adopted by all users is not unified. That is, the embodiment of the present disclosure can realize the diversification of the weights used in the fusion stage to adapt to different users. Therefore, the embodiment of the present disclosure can improve the personalization of the recommendation results when making material recommendations, so as to achieve accurate personalized recommendations for users.

[0173] In an optional example, after recommending materials to the first user according to the fusion result, the method further includes:

[0174] Determine the first user's behavior performance on the material recommendation result;

[0175] Update historical behavior data based on behavioral performance.

[0176] In an embodiment of the present disclosure, after the material recommendation result is displayed on the client used by the first user, the first user's behavior performance on the material recommendation result can be determined, and the behavior performance can be used to indicate whether the first user has viewed the listings in the material recommendation result, and which listings in the material recommendation result the first user has viewed. If the first user's recommended material behavior performance information does not exist in the historical behavior data, the information represented by the behavior performance can be added to the historical behavior data; otherwise, the first user's recommended material behavior performance information that already exists in the historical behavior data can be updated so that the updated first user's recommended material behavior performance information contains the information represented by the behavior performance.

[0177] In the embodiments of the present disclosure, by updating the historical behavior data, the information in the historical behavior data can be kept in the latest state, thereby ensuring the correctness and reliability of the weights subsequently determined based on the historical behavior data.

[0178] In an alternative example, Figure 3 As shown, the first user can send a user request to the service platform. At this time, in response to the user request, the recommendation system can use multiple recall strategies to recall materials for the first user to obtain multiple recall results (ie, recall materials through multiple recall strategies).

[0179] Specifically, if the first user is a new user, the weight corresponding to each recall strategy can be obtained according to the preset corresponding relationship stored in the Remote Dictionary Server (Redis) database, and multiple recall results can be weighted and fused based on the obtained weights; if the first user is an old user, for the recall strategy that hits the bandit fusion strategy (equivalent to the recall strategy in the first category of recall strategies mentioned above), the weight corresponding to the recall strategy of the first user calculated by the bandit can be obtained from the Redis database, and for the recall strategy that does not hit the bandit fusion strategy (equivalent to the recall strategy in the second category of recall strategies mentioned above), it can be given the maximum weight. After the weighted fusion operation is completed, the material recommendation results including the sorted multiple housing sources can be returned to the first user through the back-end application programming interface (Application Programming Interface, API). At this time, the recommendation system will record the recall strategy ID and exposure ID of each housing source in the material recommendation result, as well as the scenario information of the recommended business scenario currently in which the first user is located. The information recorded by the recommendation system will flow into the back-end Kafka (i.e., Kafka) log.

[0180] After the first user receives the material recommendation result and operates at least part of the house sources in the material recommendation result (for example, clicks on the house source card used to represent the house source), the corresponding exposure and click information will flow into the Kafka log through the front-end log (for example, the front-end dig log). In this way, the back-end Kafka log can be parsed in real time to obtain the corresponding scene information, recall strategy ID and exposure ID, and then stored in the Redis database; then the exposure ID of each house source in the front-end dig log and the user exposure click and other information can be parsed in real time, combined with the user exposure click and other information obtained based on the front-end dig log, and the scene information, recall strategy ID and exposure ID obtained based on the back-end Kafka log, the click and exposure of each recall strategy of each user can be calculated, and then the click and exposure of each recall strategy of each user can be updated (equivalent to updating the material viewing times V and material exposure times C in the above text). After that, the weight calculation can be performed using the following formula to obtain the latest weight corresponding to each recall strategy of each user.

[0181]

[0182] In summary, the embodiments of the present disclosure optimize the fusion stage of the recommendation system, and personalize and update the weights corresponding to each recall strategy of the user in the fusion stage in real time based on the bandit algorithm. Specifically, each user's feedback behavior on the house will be included in the behavior of the recall strategy to which it belongs, and the weight update calculation will be performed based on the user's behavioral feedback on each recall strategy, so that the updated weight can be used as the new weight of the recall strategy in the next recommendation for weighted fusion, so that the recall strategy with good effect can be fully utilized next time, and the weights determined manually a priori can be used for new users, and the new recall strategy that is not hit by old users can be given the maximum weight, thereby improving the exploratory nature of the recommendation system. The embodiments of the present disclosure can also mine the user's own preferences in real time to improve the real-time nature and recommendation accuracy of the recommendation system.

[0183] Any recommendation method provided in the embodiments of the present disclosure may be executed by any appropriate device with data processing capabilities, including but not limited to: a terminal device and a server. Alternatively, any recommendation method provided in the embodiments of the present disclosure may be executed by a processor, such as the processor executing any recommendation method mentioned in the embodiments of the present disclosure by calling corresponding instructions stored in a memory. This will not be described in detail below.

[0184] Exemplary Devices

[0185] Figure 4 is a structural diagram of a recommendation device provided by an exemplary embodiment of the present disclosure, Figure 4The device shown includes a first acquisition module 401 , a second acquisition module 402 , a first determination module 403 , a third acquisition module 404 and a recommendation module 405 .

[0186] A first acquisition module 401 is used to adopt multiple recall strategies to recall materials for a first user to obtain multiple recall results;

[0187] The second acquisition module 402 is used to determine whether there is recommended material behavior performance information of the first user in the historical behavior data to obtain a determination result;

[0188] A first determination module 403 is used to determine a plurality of weights corresponding to a plurality of recall strategies for the first user in a corresponding weight determination manner according to the judgment result;

[0189] The third acquisition module 404 is used to fuse multiple recall results using multiple weights to obtain a fusion result;

[0190] The recommendation module 405 is used to recommend materials to the first user according to the fusion result.

[0191] In an alternative example, Figure 5 As shown, the first determination module 403 includes:

[0192] The first determination submodule 4031 is used to determine a plurality of weights corresponding to a plurality of recall strategies for the first user according to the recommended material behavior performance information of the first user when the judgment result indicates that the recommended material behavior performance information of the first user exists in the historical behavior data;

[0193] The second determination submodule 4032 is used to determine multiple weights corresponding to multiple recall strategies for the first user according to a preset correspondence relationship when the judgment result indicates that there is no recommended material behavior performance information of the first user in the historical behavior data; wherein the preset correspondence relationship is the correspondence between the recall strategy and the weight.

[0194] In an optional example, the first determining submodule 4031 includes:

[0195] A first determining unit is used to determine a first type of recall strategy among multiple recall strategies; wherein the recommended material behavior performance information of the first user includes recommended material behavior performance information related to each recall strategy in the first type of recall strategy;

[0196] A second determination unit is used to determine the number of material exposures and the number of material views corresponding to each recall strategy in the first category of recall strategies according to the recommended material behavior performance information related to each recall strategy in the first category of recall strategies;

[0197] The third determination unit is used to determine the weight corresponding to each recall strategy in the first category according to the number of material exposures and the number of material views corresponding to each recall strategy in the first category.

[0198] In an optional example, the third determining unit includes:

[0199] A first calculation subunit is used to calculate the sum of the material exposure times corresponding to various recall strategies in the first type of recall strategy;

[0200] A second calculation subunit is used to calculate the ratio of the sum to the number of material exposures corresponding to any recall strategy in the first category of recall strategies;

[0201] The third calculation subunit is used to use the formula Calculate the weight corresponding to any recall strategy; where Q is the weight corresponding to any recall strategy, V is the number of material views corresponding to any recall strategy, and S is the preset value. Is a ratio.

[0202] In an optional example, the first determining submodule 4031 includes:

[0203] a fourth determination unit, configured to determine a first type of recall strategy and a second type of recall strategy among the plurality of recall strategies; wherein the recommended material behavior performance information of the first user includes the recommended material behavior performance information related to each recall strategy in the first type of recall strategy, and the recommended material behavior performance information of the first user does not include the recommended material behavior performance information related to each recall strategy in the second type of recall strategy;

[0204] A fifth determining unit, configured to determine a weight corresponding to each recall strategy in the first category of recall strategies according to the recommended material behavior performance information related to each recall strategy in the first category of recall strategies;

[0205] The sixth determining unit is used to determine the weight corresponding to each recall strategy in the second category of recall strategies according to the weight corresponding to each recall strategy in the first category of recall strategies.

[0206] In an optional example, the sixth determining unit includes:

[0207] A screening subunit, used for screening the weight with the largest value from the weights corresponding to various recall strategies in the first category of recall strategies;

[0208] The determination subunit is used to determine the weight corresponding to each recall strategy in the second type of recall strategies; wherein each determined weight is greater than the weight with the largest value.

[0209] In an alternative example,

[0210] The first determining submodule 4031 includes:

[0211] A seventh determining unit, configured to determine a current recommended service scenario for the first user;

[0212] An eighth determination unit, configured to determine, for the first user, a plurality of weights corresponding to a plurality of recall strategies according to the recommended material behavior performance information of the first user related to the current recommended business scenario;

[0213] and / or,

[0214] The second determining submodule 4032 includes:

[0215] A ninth determining unit, configured to determine a current recommended service scenario for the first user;

[0216] The tenth determining unit is used to determine a plurality of weights corresponding to a plurality of recall strategies for the first user according to a preset corresponding relationship corresponding to a current recommendation business scenario.

[0217] In an optional example, the device further includes:

[0218] A second determination module is used to determine the behavior of the first user on the material recommendation result after recommending the material to the first user according to the fusion result;

[0219] The update module is used to update the historical behavior data according to the behavior performance.

[0220] Exemplary Electronic Devices

[0221] Below, reference Figure 6 The electronic device according to the embodiment of the present disclosure is described. The electronic device may be any one or both of the first device and the second device, or a stand-alone device independent of them, and the stand-alone device may communicate with the first device and the second device to receive the collected input signals from them.

[0222] Figure 6 A block diagram of an electronic device 60 according to an embodiment of the present disclosure is illustrated.

[0223] like Figure 6 As shown, the electronic device 60 includes one or more processors 61 and a memory 62 .

[0224] The processor 61 may be a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 60 to perform desired functions.

[0225] The memory 62 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, a random access memory (RAM) and / or a cache memory (cache), etc. The non-volatile memory may include, for example, a read-only memory (ROM), a hard disk, a flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 61 may run the program instructions to implement the recommended methods of the various embodiments of the present disclosure described above and / or other desired functions. Various contents such as input signals, signal components, noise components, etc. may also be stored in the computer-readable storage medium.

[0226] In one example, the electronic device 60 may further include: an input device 63 and an output device 64, and these components are interconnected via a bus system and / or other forms of connection mechanisms (not shown).

[0227] For example, when the electronic device 60 is the first device or the second device, the input device 63 may be a microphone or a microphone array. When the electronic device 60 is a stand-alone device, the input device 63 may be a communication network connector for receiving collected input signals from the first device and the second device.

[0228] In addition, the input device 63 may also include, for example, a keyboard, a mouse, etc.

[0229] The output device 64 can output various information to the outside, and can include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto.

[0230] Of course, to simplify, Figure 6 Only some of the components related to the present disclosure in the electronic device 60 are shown, and components such as a bus, an input / output interface, etc. are omitted. In addition, according to specific application situations, the electronic device 60 may also include any other appropriate components.

[0231] Exemplary computer program products and computer-readable storage media

[0232] In addition to the above-mentioned methods and devices, an embodiment of the present disclosure may also be a computer program product, which includes computer program instructions, which, when executed by a processor, enable the processor to execute the steps of the recommended method according to various embodiments of the present disclosure described in the above-mentioned "Exemplary Method" section of this specification.

[0233] The computer program product may be written in any combination of one or more programming languages ​​to write program code for performing the operations of the disclosed embodiments, including object-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0234] In addition, an embodiment of the present disclosure may also be a computer-readable storage medium having computer program instructions stored thereon, which, when executed by a processor, causes the processor to execute the steps of the recommended method according to various embodiments of the present disclosure described in the above “Exemplary Method” section of this specification.

[0235] The computer readable storage medium can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can include, for example, but is not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, 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 disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0236] The basic principles of the present disclosure are described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, effects, etc. mentioned in the present disclosure are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. are required by each embodiment of the present disclosure. In addition, the specific details disclosed above are only for the purpose of illustration and ease of understanding, and are not limitations. The above details do not limit the present disclosure to the necessity of adopting the above specific details to be implemented.

[0237] Each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system embodiment, since it basically corresponds to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0238] The block diagrams of the devices, apparatuses, equipment, and systems involved in this disclosure are only illustrative examples and are not intended to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagrams. As will be appreciated by those skilled in the art, these devices, apparatuses, equipment, and systems can be connected, arranged, and configured in any manner. Words such as "including," "comprising," "having," and the like are open words, referring to "including but not limited to," and can be used interchangeably therewith. The words "or" and "and" used herein refer to the words "and / or," and can be used interchangeably therewith, unless the context clearly indicates otherwise. The word "such as" used herein refers to the phrase "such as but not limited to," and can be used interchangeably therewith.

[0239] The method and apparatus of the present disclosure may be implemented in many ways. For example, the method and apparatus of the present disclosure may be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above order of steps for the method is for illustration only, and the steps of the method of the present disclosure are not limited to the order specifically described above, unless otherwise specifically stated. In addition, in some embodiments, the present disclosure may also be implemented as a program recorded in a recording medium, which includes machine-readable instructions for implementing the method according to the present disclosure. Therefore, the present disclosure also covers a recording medium storing a program for executing the method according to the present disclosure.

[0240] It should also be noted that in the apparatus, device and method of the present disclosure, each component or each step can be decomposed and / or recombined. Such decomposition and / or recombination should be regarded as equivalent solutions of the present disclosure.

[0241] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present disclosure. Therefore, the present disclosure is not intended to be limited to the aspects shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

[0242] The above description has been given for the purpose of illustration and description. In addition, this description is not intended to limit the embodiments of the present disclosure to the forms disclosed herein. Although multiple example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, changes, additions and sub-combinations thereof.

Claims

1. A recommendation method, characterized in that: include: Use multiple recall strategies to recall materials for the first user to obtain multiple recall results; Determine whether there is recommended material behavior performance information of the first user in the historical behavior data to obtain a determination result; According to the judgment result, determining a plurality of weights corresponding to the plurality of recall strategies for the first user in a corresponding weight determination manner; Using the multiple weights, fusing the multiple recall results to obtain a fusion result; Recommending materials to the first user according to the fusion result; Wherein, determining multiple weights corresponding to the multiple recall strategies for the first user in a corresponding weight determination manner according to the judgment result includes: In a case where the judgment result indicates that the recommended material behavior performance information of the first user exists in the historical behavior data, determining multiple weights corresponding to the multiple recall strategies for the first user according to the recommended material behavior performance information of the first user, including: Determine a first category of recall strategies among the multiple recall strategies; the recommended material behavior performance information of the first user includes recommended material behavior performance information related to each recall strategy in the first category of recall strategies; determine the number of material exposures and the number of material views corresponding to each recall strategy in the first category of recall strategies based on the recommended material behavior performance information related to each recall strategy in the first category of recall strategies; determine the weight corresponding to each recall strategy in the first category of recall strategies based on the number of material exposures and the number of material views corresponding to each recall strategy in the first category of recall strategies.

2. The method according to claim 1, characterized in that The determining, according to the judgment result, a plurality of weights corresponding to the plurality of recall strategies for the first user in a corresponding weight determination manner includes: When the judgment result indicates that there is no recommended material behavior performance information of the first user in the historical behavior data, multiple weights corresponding to the multiple recall strategies are determined for the first user according to a preset corresponding relationship; wherein the preset corresponding relationship is a corresponding relationship between the recall strategy and the weight.

3. The method according to claim 1, characterized in that The determining, according to the number of material exposures and the number of material viewings corresponding to each recall strategy in the first category of recall strategies, a weight corresponding to each recall strategy in the first category of recall strategies includes: Calculating the sum of the material exposure times corresponding to various recall strategies in the first type of recall strategies; Calculating a ratio of the sum to the number of material exposures corresponding to any recall strategy in the first category of recall strategies; Using the formula Calculate the weight corresponding to any of the recall strategies; wherein Q is the weight corresponding to any of the recall strategies, V is the number of material views corresponding to any of the recall strategies, and S is a preset value. is the ratio.

4. The method according to claim 1, characterized in that: The determining, for the first user, a plurality of weights corresponding to the plurality of recall strategies according to the recommended material behavior performance information of the first user includes: Determine a first type of recall strategy and a second type of recall strategy among the multiple recall strategies; wherein the recommended material behavior performance information of the first user includes the recommended material behavior performance information related to each recall strategy in the first type of recall strategy, and the recommended material behavior performance information of the first user does not include the recommended material behavior performance information related to each recall strategy in the second type of recall strategy; Determining a weight corresponding to each recall strategy in the first category of recall strategies according to the recommended material behavior performance information related to each recall strategy in the first category of recall strategies; According to the weight corresponding to each recall strategy in the first category of recall strategies, the weight corresponding to each recall strategy in the second category of recall strategies is determined.

5. The method according to claim 4, characterized in that The determining, according to the weight corresponding to each recall strategy in the first category of recall strategies, the weight corresponding to each recall strategy in the second category of recall strategies includes: Selecting the weight with the largest value from the weights corresponding to the various recall strategies in the first category of recall strategies; Determine the weight corresponding to each recall strategy in the second type of recall strategies; wherein each determined weight is greater than the weight with the largest value.

6. The method according to claim 2, characterized in that The step of determining, for the first user, a plurality of weights corresponding to the plurality of recall strategies based on the recommended material behavior performance information of the first user includes: Determining a current recommended business scenario for the first user; Determining, for the first user, a plurality of weights corresponding to the plurality of recall strategies according to the recommended material behavior performance information of the first user related to the current recommended business scenario; and / or, The step of determining, for the first user, a plurality of weights corresponding to the plurality of recall strategies according to the preset corresponding relationship includes: Determining a current recommended business scenario for the first user; According to the preset corresponding relationship corresponding to the current recommended business scenario, multiple weights corresponding to the multiple recall strategies are determined for the first user.

7. The method according to claim 1, characterized in that After recommending materials to the first user according to the fusion result, the method further includes: Determining the first user's behavior performance on the material recommendation result; The historical behavior data is updated according to the behavior performance.

8. A recommendation device, characterized in that: include: A first acquisition module, used to adopt multiple recall strategies to recall materials for the first user to obtain multiple recall results; A second acquisition module is used to determine whether there is the recommended material behavior performance information of the first user in the historical behavior data to obtain a determination result; A first determination module, configured to determine, according to the judgment result and in a corresponding weight determination manner, a plurality of weights corresponding to the plurality of recall strategies for the first user; A third acquisition module, configured to fuse the multiple recall results using the multiple weights to obtain a fusion result; A recommendation module, configured to recommend materials to the first user according to the fusion result; Wherein, the first determining module includes: A first determination submodule is configured to determine, when the judgment result indicates that there is recommended material behavior performance information of the first user in the historical behavior data, a plurality of weights corresponding to the plurality of recall strategies for the first user according to the recommended material behavior performance information of the first user; The first determining submodule includes: A first determining unit is used to determine a first type of recall strategy among the multiple recall strategies; wherein the recommended material behavior performance information of the first user includes recommended material behavior performance information related to each recall strategy in the first type of recall strategy; A second determination unit is used to determine the number of material exposures and the number of material views corresponding to each recall strategy in the first category of recall strategies according to the recommended material behavior performance information related to each recall strategy in the first category of recall strategies; The third determination unit is used to determine the weight corresponding to each recall strategy in the first category of recall strategies according to the number of material exposures and the number of material views corresponding to each recall strategy in the first category of recall strategies.

9. The device according to claim 8, characterized in that The first determining module includes: The second determination submodule is used to determine multiple weights corresponding to the multiple recall strategies for the first user according to a preset correspondence relationship when the judgment result indicates that there is no recommended material behavior performance information of the first user in the historical behavior data; wherein the preset correspondence relationship is a correspondence relationship between the recall strategy and the weight.

10. The device according to claim 8, characterized in that The third determining unit includes: A first calculation subunit, used to calculate the sum of material exposure times corresponding to various recall strategies in the first type of recall strategies; A second calculation subunit is used to calculate the ratio of the sum to the number of material exposures corresponding to any recall strategy in the first type of recall strategies; The third calculation subunit is used to use the formula Calculate the weight corresponding to any of the recall strategies; wherein Q is the weight corresponding to any of the recall strategies, V is the number of material views corresponding to any of the recall strategies, and S is a preset value. is the ratio.

11. The device according to claim 9, characterized in that The first determining submodule includes: a fourth determination unit, configured to determine a first type of recall strategy and a second type of recall strategy among the multiple recall strategies; wherein the recommended material behavior performance information of the first user includes the recommended material behavior performance information related to each recall strategy in the first type of recall strategy, and the recommended material behavior performance information of the first user does not include the recommended material behavior performance information related to each recall strategy in the second type of recall strategy; a fifth determining unit, configured to determine a weight corresponding to each recall strategy in the first category of recall strategies according to the recommended material behavior performance information related to each recall strategy in the first category of recall strategies; The sixth determining unit is used to determine the weight corresponding to each recall strategy in the second category of recall strategies according to the weight corresponding to each recall strategy in the first category of recall strategies.

12. The device according to claim 11, characterized in that The sixth determining unit includes: A screening subunit, configured to screen the weight with the largest value from the weights corresponding to the various recall strategies in the first category of recall strategies; A determination subunit is used to determine the weight corresponding to each recall strategy in the second type of recall strategy; wherein each determined weight is greater than the weight with the largest value.

13. The device according to claim 9, characterized in that The first determining submodule includes: A seventh determining unit, configured to determine a current recommended service scenario for the first user; An eighth determining unit, configured to determine, for the first user, a plurality of weights corresponding to the plurality of recall strategies according to the recommended material behavior performance information of the first user related to the current recommended business scenario; and / or, The second determining submodule includes: A ninth determining unit, configured to determine a current recommended service scenario for the first user; The tenth determining unit is used to determine a plurality of weights corresponding to the plurality of recall strategies for the first user according to a preset corresponding relationship corresponding to the current recommended business scenario.

14. The device according to claim 8, characterized in that The device also includes: A second determination module is used to determine the behavior performance of the first user on the material recommendation result after recommending the material to the first user according to the fusion result; An updating module is used to update the historical behavior data according to the behavior performance.

15. A computer-readable storage medium storing a computer program, characterized in that: The computer program is used to execute the recommendation method according to any one of claims 1 to 7.

16. An electronic device, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is used to read the executable instructions from the memory and execute the instructions to implement the recommendation method according to any one of claims 1 to 7.

Images