Information processing method, information processing device, storage medium and electronic device

Abstract

The present disclosure provides an information processing method, an information processing device, a computer-readable storage medium and an electronic device, and belongs to the field of computer technology. The method includes: obtaining the historical behavior data of the target user and the object data of the object to be recommended; determining the data processing strategy of the historical behavior data and the object data according to the data volume of the historical behavior data; determining the decay time of the object to be recommended, and performing decay processing on the historical behavior data and the object data according to the decay time to generate the target feature data of the target user and the object to be recommended; processing the target feature data according to the data processing strategy to determine the object to be recommended that matches the target user. The present disclosure can improve the accuracy of information recommendation.

Description

Background Art

[0002] In order to improve the efficiency of users in obtaining effective information, various Internet companies are committed to pushing information that may be of interest to users through recommendation methods.

[0003] There are two main existing recommendation methods. One is to find other users similar to the user by analyzing the user's historical behavior data, and then recommend information based on the information that similar users are interested in. The other is to find similar information and recommend it based on the characteristics of the information itself, such as category, price, etc. These two methods recommend information from the perspective of the user and the perspective of information respectively. However, due to the relatively single dimension of their feature data, the accuracy of information recommendation is not high, and when the amount of historical behavior data is small, the accuracy of the algorithm will be greatly affected.

[0004] Therefore, it is necessary to provide a method that can effectively improve the recommendation accuracy.

[0005] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0006] The present disclosure provides an information processing method, an information processing device, a computer-readable storage medium and an electronic device, thereby improving the problem of low accuracy of information recommendation in the prior art at least to a certain extent.

[0007] Other features and advantages of the present disclosure will become apparent from the following detailed description, or may be learned in part by the practice of the present disclosure.

[0008] According to a first aspect of the present disclosure, there is provided an information processing method, the method comprising: obtaining historical behavior data of a target user and object data of an object to be recommended; determining a data processing strategy for the historical behavior data and the object data according to the data volume of the historical behavior data; determining a decay time of the object to be recommended, and performing decay processing on the historical behavior data and the object data according to the decay time to generate target feature data of the target user and the object to be recommended; processing the target feature data according to the data processing strategy to determine an object to be recommended that matches the target user; wherein the decay time is used to represent the time interval between the generation time of the initial recommendation score of the object to be recommended and the current time.

[0009] In an exemplary embodiment of the present disclosure, determining the data processing strategy for the historical behavior data and the object data based on the data volume of the historical behavior data includes: when the data volume of the historical behavior data is greater than a data volume threshold, determining that the data processing strategy for the historical behavior data and the object data is a first processing strategy; when the data volume of the historical behavior data is not greater than the data volume threshold, determining that the data processing strategy for the historical behavior data and the object data is a second processing strategy.

[0010] In an exemplary embodiment of the present disclosure, the attenuation time of the object to be recommended is determined, and the historical behavior data and the object data are attenuated according to the attenuation time to generate target feature data of the target user and the object to be recommended, including: determining an initial recommendation score of the object to be recommended according to the historical behavior data or the object data; reducing the initial recommendation score according to the attenuation time to generate the target feature data.

[0011] In an exemplary embodiment of the present disclosure, the step of reducing the initial recommendation score according to the decay time to generate the target feature data includes: constructing a decay function in which the current recommendation score and the decay time are inversely proportional, and reducing the initial recommendation score by the decay function to obtain the current recommendation score of the object to be recommended; and processing the historical behavior data or the object data according to the current recommendation score of the object to be recommended to generate the target feature data.

[0012] In an exemplary embodiment of the present disclosure, reducing the initial recommendation score by the attenuation function to obtain the current recommendation score of the object to be recommended includes: calculating the attenuation weight of the initial recommendation score by the attenuation function; and multiplying the attenuation weight by the initial recommendation score to obtain the current recommendation score of the object to be recommended.

[0013] In an exemplary embodiment of the present disclosure, calculating the attenuation weight of the initial recommendation score by using the attenuation function includes: calculating the attenuation weight of the initial recommendation score by using the following formula:

[0014] y=ae -μΔt

[0015] Among them, △t is the decay time, a is the initial recommendation score, and μ is the decay rate.

[0016] In an exemplary embodiment of the present disclosure, the target feature data is processed according to the data processing strategy to determine the object to be recommended that matches the target user, including: processing the target feature data according to the first processing strategy to obtain similar users of the target user, and determining the object to be recommended that matches the target user based on the association scores between the similar users and the object to be recommended; and / or processing the target feature data according to the second processing strategy to determine the object to be recommended that matches the target user.

[0017] In an exemplary embodiment of the present disclosure, the target feature data is processed according to the first processing strategy to obtain similar users of the target user, and the object to be recommended that matches the target user is determined based on the association scores between the similar users and the object to be recommended, including: generating a behavior feature matrix of the target user based on the target feature data, and determining similar users of the target user based on the behavior feature matrix; determining the target user's interest in the object to be recommended based on the similarity between the similar users and the target user, and the association scores of the similar users to the object to be recommended, and determining the object to be recommended that matches the target user based on the interest.

[0018] In an exemplary embodiment of the present disclosure, the target feature data is processed according to the second processing strategy to determine the object to be recommended that matches the target user, including: generating a user feature vector of the target user and an object feature vector of the object to be recommended according to the target feature data; and determining the object to be recommended that matches the target user by calculating the similarity between the user feature vector and the object feature vector.

[0019] In an exemplary embodiment of the present disclosure, determining the object to be recommended that matches the target user by calculating the similarity between the user feature vector and the object feature vector includes: determining similar objects to the object to be recommended based on the historical behavior data; and determining the object to be recommended that matches the target user based on the association score between the target user and the similar objects.

[0020] According to a second aspect of the present disclosure, an information processing device is provided, comprising: an acquisition module for acquiring historical behavior data of a target user and object data of an object to be recommended; a determination module for determining a data processing strategy for the historical behavior data and the object data according to the data volume of the historical behavior data; a generation module for determining a decay time of the object to be recommended, and performing decay processing on the historical behavior data and the object data according to the decay time to generate target feature data of the target user and the object to be recommended; a processing module for processing the target feature data according to the data processing strategy to determine an object to be recommended that matches the target user; wherein the decay time is used to represent the time interval between the generation time of the initial recommendation score of the object to be recommended and the current time.

[0021] In an exemplary embodiment of the present disclosure, the determination module is used to determine that the data processing strategy of the historical behavior data and the object data is a first processing strategy when the data volume of the historical behavior data is greater than a data volume threshold, and to determine that the data processing strategy of the historical behavior data and the object data is a second processing strategy when the data volume of the historical behavior data is not greater than the data volume threshold.

[0022] In an exemplary embodiment of the present disclosure, the generation module is used to determine the initial recommendation score of the object to be recommended according to the historical behavior data or the object data, reduce the initial recommendation score according to the decay time, and generate the target feature data.

[0023] In an exemplary embodiment of the present disclosure, the generation module is also used to construct a decay function in which the current recommendation score is inversely proportional to the decay time, and reduce the initial recommendation score through the decay function to obtain the current recommendation score of the object to be recommended, and process the historical behavior data or the object data according to the current recommendation score of the object to be recommended to generate the target feature data.

[0024] In an exemplary embodiment of the present disclosure, the generating module is further configured to calculate a decay weight of the initial recommendation score by using the decay function, and multiply the decay weight by the initial recommendation score to obtain a current recommendation score of the object to be recommended.

[0025] In an exemplary embodiment of the present disclosure, the generating module is further configured to calculate the decay weight of the initial recommendation score by the following formula:

[0026] y=ae -μΔt

[0027] Among them, △t is the decay time, a is the initial recommendation score, and μ is the decay rate.

[0028] In an exemplary embodiment of the present disclosure, the processing module is used to process the target feature data according to the first processing strategy to obtain similar users of the target user, and determine the object to be recommended that matches the target user based on the association score between the similar users and the object to be recommended, and / or process the target feature data according to the second processing strategy to determine the object to be recommended that matches the target user.

[0029] In an exemplary embodiment of the present disclosure, the processing module is also used to generate a behavior characteristic matrix of the target user based on the target characteristic data, and determine similar users of the target user based on the behavior characteristic matrix, determine the target user's interest in the object to be recommended based on the similarity between the similar users and the target user, and the association score of the similar users to the object to be recommended, and determine the object to be recommended that matches the target user based on the interest.

[0030] In an exemplary embodiment of the present disclosure, the processing module is also used to generate a user feature vector of the target user and an object feature vector of the object to be recommended according to the target feature data, and determine the object to be recommended that matches the target user by calculating the similarity between the user feature vector and the object feature vector.

[0031] In an exemplary embodiment of the present disclosure, the processing module is further used to determine similar objects of the object to be recommended based on the historical behavior data, and determine the object to be recommended that matches the target user based on the association score between the target user and the similar objects.

[0032] According to a third aspect of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, any one of the above-mentioned information processing methods is implemented.

[0033] According to a fourth aspect of the present disclosure, an electronic device is provided, comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to perform any one of the above-mentioned information processing methods by executing the executable instructions.

[0034] The present disclosure has the following beneficial effects:

[0035] According to the information processing method, information processing device, computer-readable storage medium and electronic device in this exemplary embodiment, the data processing strategy of historical behavior data and object data can be determined according to the data volume of historical behavior data, the decay time of the object to be recommended can be determined, and the historical behavior data and object data can be decayed according to the decay time to generate target feature data of the target user and the object to be recommended, and then the target feature data can be processed according to the data processing strategy to determine the object to be recommended that matches the target user. On the one hand, by determining the decay time of the object to be recommended and decaying the historical behavior data and object data according to the decay time, the dynamic change amount of the user's interest can be determined in combination with the influence of time factors on the user's interest, thereby improving the accuracy of information recommendation and reducing the time it takes for users to find objects of interest; on the other hand, by determining the data processing strategy of historical behavior data and object data, the data processing method can be predetermined to avoid the problem of low information recommendation accuracy caused by using the same method, while also improving the efficiency of data processing.

[0036] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The drawings herein are incorporated into the specification and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and are used together with the specification to explain the principles of the present disclosure. Obviously, the drawings described below are only some embodiments of the present disclosure, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0038] Figure 1 A flowchart showing an information processing method in this exemplary embodiment;

[0039] Figure 2 A data processing flow chart showing a first processing strategy in this exemplary embodiment;

[0040] Figure 3 A sub-flow chart showing an information processing method in this exemplary embodiment;

[0041] Figure 4 A schematic diagram showing an interactive behavior in this exemplary embodiment;

[0042] Figure 5 A sub-flow chart showing another information processing method in this exemplary embodiment;

[0043] Figure 6 A sub-flow chart showing another information processing method in this exemplary embodiment;

[0044] Figure 7 A schematic diagram showing a behavior characteristic matrix in this exemplary embodiment;

[0045] Figure 8 A flow chart of determining the interest level of a target user in this exemplary embodiment is shown;

[0046] Fig. 9 A flow chart of generating a feature vector in this exemplary embodiment is shown;

[0047] Fig.10 Another flow chart for generating a feature vector in this exemplary embodiment is shown;

[0048] Fig.11 A data processing flow chart showing a second processing strategy in this exemplary embodiment;

[0049] Fig.12 A flowchart showing another information processing method in this exemplary embodiment;

[0050] Fig.13 A structural block diagram of an information processing device according to the exemplary embodiment is shown;

[0051] Fig.14 An electronic device for implementing the above method in this exemplary embodiment is shown. DETAILED DESCRIPTION

[0052] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that the disclosure will be more comprehensive and complete and to fully convey the concepts of the example embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

[0053] In view of the above-mentioned various problems, the exemplary embodiments of the present disclosure first provide an information processing method. The method can be executed by a computer or a server in the background of an application to determine the recommended objects that match the target user, for example, in a music application, the albums, music or singers that match the target user can be determined.

[0054] Figure 1 A process of this exemplary embodiment is shown, which may include the following steps S110 to S140:

[0055] Step S110: Obtain the historical behavior data of the target user and the object data of the object to be recommended.

[0056] Among them, the historical behavior data can be log data related to the recommended object, that is, the content or object affected by the interactive behavior of the target user, such as the behavioral data of any one or more interactive behaviors such as clicks, searches, purchases, and favorites of the recommended object by the target user in the past period of time; the recommended object refers to the content or object that needs to be recommended to the user, which can include the recommended object with which the target user has interacted and the recommended object with which the target user has not interacted; the object data of the recommended object can include feature data such as the category, price, and user rating of the recommended object. In addition, the historical behavior data can include implicit data and explicit data according to the interactive relationship between the target user and the recommended object, wherein implicit data can be data generated by any one or more behaviors such as clicks and views of the recommended object by the target user, and explicit data can be data generated by any one or more behaviors such as comments, searches, purchases, sharing, and favorites of the recommended object by the target user.

[0057] In this exemplary embodiment, the historical behavior data of the target user and the object data of the object to be recommended can be read through a corresponding database or data platform. For example, the historical behavior data of the target user and the object data of the object to be recommended can be extracted from the above database or data platform using the user ID of the target user and the object ID of the object to be recommended as indexes respectively.

[0058] Furthermore, considering that the type of interaction between the user and the recommended object can also indicate the user's interest in the recommended object, in an optional implementation, reference Figure 2 As shown, after obtaining the historical behavior data of the target user through step S210, the implicit data and explicit data in the historical behavior data can also be determined through step S220. For example, the target user's interactive behavior towards the recommended object can be determined based on the historical behavior data, and then it can be determined whether the interactive behavior matches the interactive behavior corresponding to the above-mentioned implicit data and explicit data, so as to divide the historical behavior data into implicit data and explicit data.

[0059] Step S120: Determine a data processing strategy for the historical behavior data and the object data according to the data volume of the historical behavior data.

[0060] The data processing strategy refers to the method for processing the above-mentioned historical behavior data and object data, which may include the processing data, processing steps, and data formats of the historical behavior data and object data, such as the dimension of the output data.

[0061] Historical behavior data can be used as an important basis for analyzing the interest of target users. When the amount of historical behavior data is large, analyzing the historical behavior data can obtain a more accurate target user's interest, but when the amount of historical behavior data is small, the historical behavior data cannot fully reflect the target user's interest. Therefore, the data processing strategy for historical behavior data and object data can be determined based on the amount of historical behavior data. For example, the processing data ratio of historical behavior data and object data can be determined separately based on the amount of historical behavior data, as well as the processing steps of these two types of data.

[0062] In an optional implementation, step S120 may be implemented by the following method:

[0063] When the data volume of the historical behavior data is greater than the data volume threshold, the data processing strategy for the historical behavior data and the object data is determined to be the first processing strategy.

[0064] When the data volume of the historical behavior data is not greater than the data volume threshold, the data processing strategy for the historical behavior data and the object data is determined to be the second processing strategy.

[0065] The data volume threshold can be set by the operator based on experience. For example, the data volume threshold of historical behavior data within one month can be set to 100,000, and the data volume of historical behavior data within one week can be set to 10,000, based on the time range corresponding to the historical behavior data; the first processing strategy and the second processing strategy refer to the data processing strategies of historical behavior data and object data when the data volume of historical behavior data is greater than the data volume threshold and not greater than the data volume threshold, respectively, and can include the proportion of data to be processed in the historical behavior data and object data, the data processing steps, the format of output data, etc. In this exemplary embodiment, according to the size of the data volume of the historical behavior data, the first processing strategy can be to analyze the historical behavior data, determine similar users of the target user, and determine the recommended objects that match the target user based on the interest of the similar users in the recommended objects; the second processing strategy can be to analyze the historical behavior data and object data to determine the recommended objects that match the target user.

[0066] As mentioned above, the amount of historical behavior data determines whether the historical behavior data can fully reflect the interest of the target user, and the size of the data determines the data processing strategy of the historical behavior data and the object data. Therefore, the correlation between the amount of historical behavior data and the data processing strategy can be established. When the amount of historical behavior data is greater than the data volume threshold, it means that the historical behavior data can fully reflect the interest of the target user in the recommended object. Therefore, the first processing strategy can be used to process the above historical behavior data and object data; when the amount of historical behavior data is not greater than the above data volume threshold, it means that the historical behavior data cannot fully reflect the interest of the target user in the recommended object. Only analyzing the historical behavior data cannot accurately determine the interest of the target user. Therefore, the second data strategy can be used to process the above historical behavior data and object data.

[0067] By determining the corresponding data processing strategy according to the amount of historical behavior data, the data processing method of historical behavior data and object data can be determined in advance, avoiding the problem of low calculation accuracy caused by using the same data processing method, and also improving data processing efficiency.

[0068] Step S130: Determine the decay time of the object to be recommended, and perform decay processing on the above historical behavior data and object data according to the decay time to generate target feature data of the target user and the object to be recommended.

[0069] The decay time can be used to represent the time interval between the generation time of the initial recommendation score of the recommended object and the current time. For example, the generation time of the initial recommendation score of the recommended object is t p , the current time is t n , then the decay time △t=t n -t p . The initial recommendation score can be used to indicate the initial popularity of the object to be recommended and the possibility of the object to be recommended to the target user. Generally, the higher the initial recommendation score, the higher the possibility of the object to be recommended to the target user, and vice versa. The target feature data can be feature data generated by the historical behavior data of the target user and the object data of the object to be recommended, and can include the interaction behavior characteristics, interaction time, interaction location, or category characteristics and price characteristics of the object to be recommended between the target user and the object to be recommended. In addition, the target feature data can also include the target user's own feature data such as gender, age, occupation, region, etc.

[0070] The applicant of the present disclosure has discovered through research that the user's interest in the recommended object will decline over time. Therefore, the above historical behavior data and object data can be decayed according to the decay time of the recommended object to generate target feature data of the target user and the recommended object.

[0071] Generally, as time goes by, if the target user does not have any new interaction with the recommended object, it can be said that the target user's interest in the recommended object is decreasing; at the same time, during the recommendation process, the recommended object will be updated over time, that is, if the recommended object is not updated, the target user's interest in the recommended object will also decrease over time. Figure 3 As shown, step S130 can be implemented by following steps S310 to S320:

[0072] Step S310: Determine the initial recommendation score of the object to be recommended based on the above historical behavior data or object data.

[0073] Step S320: Reduce the initial recommendation score according to the decay time to generate the target feature data.

[0074] In this exemplary embodiment, the initial recommendation score of the object to be recommended can be determined based on the above-mentioned historical behavior data or object data. For example, when the target user has interacted with the object to be recommended, the type of interaction behavior and duration of the interaction of the target user with respect to the recommended object can be determined based on the historical behavior data, and then the initial recommendation score can be determined. When the target user has no interaction with the object to be recommended, the initial recommendation score of the object to be recommended can be determined based on the object data. The initial recommendation score can be a recommendation score generated in advance by an operator, or it can also be a default recommendation score for the object to be recommended.

[0075] Figure 4 A schematic diagram of the interaction behavior between a user and an object to be recommended is shown. As shown in the figure, target user 1 has generated the interaction behaviors of collecting, adding to cart and browsing for recommended objects A, B, and C, respectively; target user 2 has generated the interaction behaviors of collecting, clicking and adding to cart for recommended objects A, B, and D, respectively; target user 3 has generated the interaction behaviors of clicking and collecting for recommended objects B and D, respectively. Therefore, according to the interaction behavior type of the target user with the recommended object, the initial recommendation score can be set to the score corresponding to the above interaction behavior type. Referring to Table 1 below, according to the interaction behavior type of the target user with the object to be recommended, the initial recommendation score of the "collection" and "browsing" interaction behaviors can be determined as 5, and the initial recommendation score of the "click" and "add to cart" interaction behaviors can be determined as 3.

[0076] Table 1

[0077]

[0078]

[0079] After determining the initial recommendation score of the object to be recommended, the initial recommendation score may be reduced according to the decay time to obtain the recommendation score of each object to be recommended at the current moment, thereby generating target feature data.

[0080] In order to facilitate determination of the current recommendation score of the object to be recommended, in an optional implementation, the current recommendation score of the object to be recommended may be calculated using a pre-built decay function, that is, Figure 5 As shown, step S320 can also be implemented by following steps S510 to S520:

[0081] Step S510: construct a decay function in which the current recommendation score and the decay time are in inverse proportion, and use the decay function to reduce the initial recommendation score to obtain the current recommendation score of the object to be recommended.

[0082] Step S520: Process the historical behavior data or object data according to the current recommendation score of the object to be recommended to generate target feature data.

[0083] Among them, the attenuation function can be any function in which the current recommendation score and the attenuation time are inversely proportional. For example, the attenuation function can be set to S=-k△t+b, where S is the current recommendation score, △t is the attenuation time, and k and b are constants; the current recommendation score is the recommendation score of the object to be recommended at the current moment. Generally, if the target user has not interacted with the object to be recommended within the attenuation time, the current recommendation score should be less than the initial recommendation score.

[0084] By performing decay calculation on the initial recommendation score according to a pre-constructed decay function, the current recommendation score of the object to be recommended after time decay can be obtained, and then the above-mentioned historical behavior data and object data can be processed according to the current recommendation score of the object to be recommended. For example, the initial recommendation score can be added to the historical behavior data and object data to generate target feature data of the target user and the object to be recommended.

[0085] In an optional implementation, the above-mentioned attenuation function can also be set as a function of calculating the attenuation factor of the initial recommendation score, and the attenuation factor can be used to indicate the degree of attenuation of the initial recommendation score. Therefore, the above-mentioned method of reducing the initial recommendation score by the attenuation function to obtain the current recommendation score of the object to be recommended can be implemented by the following method:

[0086] The decay weight of the initial recommendation score is calculated by the decay function.

[0087] The above decay weight is multiplied by the initial recommendation score to obtain the current recommendation score of the object to be recommended.

[0088] The decay weight may be used to indicate the decay degree of the initial recommendation score. The higher the decay weight, the higher the decay degree of the initial recommendation score. Conversely, the lower the decay degree of the initial recommendation score.

[0089] By calculating the decay weight of the initial recommendation score through the decay function, the accuracy of determining the current recommendation score can be improved. By multiplying the decay weight with the initial recommendation score, the current recommendation score of the recommended object at the current moment can be obtained. For example, assuming that the initial recommendation score of the recommended object i is M i , the decay weight of the initial recommendation score is y i , then the current recommendation score is M i *y i .

[0090] The applicant of the present disclosure has found through research that the user's interest in the recommended object is related to the user's memory of the recommended object. According to people's memory rules, the user's memory of the recommended object shows a forgetting rule as shown in Table 2 below:

[0091] Table 2

[0092] Time interval Memory Just completed 100％ 20 minutes later 58.2％ 1 hour later 44.2％ 8-9 hours 35.8％ 1 day 33.7％ 2 days 27.8％ 6 days 25.4％ 30 days 21.1％

[0093] By fitting the above memory curve, the change law of the target user's memory of the recommended object can be determined. That is, the decay weight of the initial recommendation score of the recommended object conforms to the memory law of the following formula (1):

[0094] y=ae -μt (1)

[0095] Among them, y is the decay weight of the initial recommendation score of the recommended object, a is the initial recommendation score, μ is the decay rate, and t is the time interval.

[0096] Therefore, in an optional implementation, when calculating the decay weight of the initial recommendation score by using the decay function, the decay weight can be calculated by the following formula (2):

[0097] y=ae -μΔt (2)

[0098] Among them, △t is the decay time, a is the initial recommendation score, and μ is the decay rate.

[0099] After obtaining the current recommendation score of the object to be recommended, the above historical behavior data or object data can be processed according to the current recommendation score to obtain the target feature data. For example, when the target user has an interactive behavior with the object to be recommended, the initial recommendation score determined according to the interactive behavior type can be attenuated, and the current recommendation score obtained can be used as feature data and added to the behavior feature data generated according to the historical behavior data to obtain the target feature data; or, when the target user has no interactive behavior with the object to be recommended, the initial recommendation score determined according to the object data can be attenuated, and the current recommendation score obtained can be used as feature data and added to the feature data generated according to the historical behavior data and the object data to obtain the target feature data.

[0100] Step S140: Process the target feature data according to the data processing strategy to determine the recommended object that matches the target user.

[0101] After the target feature data is obtained, the target feature data may be processed according to the above data processing strategy to determine the object to be recommended that matches the target user.

[0102] In an optional embodiment, reference Figure 6 As shown, step S140 can be obtained by following the steps S610 to S620:

[0103] Step S610: Process the target feature data according to the first processing strategy to obtain similar users of the target user, and determine the recommended object that matches the target user based on the relevance score between the similar users and the recommended object.

[0104] Step S620: Process the target feature data according to the second processing strategy to determine the recommended object that matches the target user.

[0105] Among them, similar users may be users who have the same or similar preferences as the target user; the relevance scores between similar users and the objects to be recommended may be data such as the ratings given by similar users to the recommended objects.

[0106] In an optional implementation, step S610 may be implemented by the following method:

[0107] A behavior feature matrix of the target user is generated according to the target feature data, and similar users of the target user are determined according to the behavior feature matrix.

[0108] According to the similarity between the similar users and the target user, and the relevance score of the similar users to the recommended object, the interest of the target user in the recommended object is determined, and the recommended object matching the target user is determined according to the interest.

[0109] like Figure 2 As shown in step S230, the interaction behavior between the target user and the recommended object can be determined based on the target feature data, and a behavior feature matrix of the target user can be generated. Taking the above target users 1, 2, and 3 as an example, the initial recommendation score can be multiplied by the decay weight to generate the following: Figure 7 The behavioral feature matrix shown in Figure 1 is ij represents the attenuation weight of the target user i for the recommended object j. Then, step S240 can be executed, that is, the user set similar to the target user is calculated through the behavior feature matrix of the target user, that is, the similar users of the target user. For example, the similarity between the feature data of the target user can be calculated through the corresponding similarity calculation formula to determine the similar users. Taking cosine similarity as an example, the similarity between any two target users can be calculated through the following formula (3):

[0110]

[0111] Among them, A and B are feature vectors generated by feature data of different target users respectively.

[0112] After calculating the similarity between any two target users, other target users whose similarity is greater than a certain threshold can be determined as similar users of the target user according to the similarity. Alternatively, Figure 2 As shown, the similarities can also be sorted in descending order according to step S250, and other target users corresponding to the first k similarities are determined as similar users of the target user, where k is a positive integer. Then, the interest of the target user in the recommended object can be determined based on the similarity between the similar users and the target user, and the relevance score of the similar users to the recommended object, and the recommended object matching the target user can be determined based on the interest. For example, the interest can be sorted, so that the recommended objects corresponding to the first m interest are determined as the recommended objects matching the target user, where m is a positive integer.

[0113] For example, refer to Figure 8 As shown, for target user 1, the relevance scores of similar users 1, 2 and 3 for the recommended object A are r11, r12 and r13 respectively, and the similarities between target user 1 and similar users 1, 2 and 3 are w11, w12 and w13 respectively. Therefore, in an optional implementation, the interest of the target user in the recommended object can be obtained as follows:

[0114] P(u,j)=∑ v∈S(q,K)∩N(j) w uv *r vj (4)

[0115] Where N(j) represents the set of similar users of the recommended object j, rvj represents the relevance score of the target user V to the recommended object j. uv Represents the similarity between the target user u and similar user v.

[0116] In an optional embodiment, when determining the target user's interest in the recommended object, it can also be determined directly based on the relevance scores of similar users towards the recommended object. For example, the average of the relevance scores of similar users towards the recommended object can be directly determined as the target user's interest in the recommended object.

[0117] When the target feature data is processed according to the second processing strategy, in an optional implementation, the target user's interest in the recommended object can be determined by determining the matching degree between the target user's interests and the features of the recommended object. Specifically, the processing of the target feature data according to the second processing strategy in step S620 can be implemented by the following method:

[0118] Generate a user feature vector of the target user and an object feature vector of the object to be recommended respectively according to the target feature data;

[0119] By calculating the similarity between the user feature vector and the object feature vector, the recommended object matching the target user is determined.

[0120] Among them, the user feature vector can be used to represent the characteristics of the recommended objects preferred by the target user, which may include the category, price, etc. of the recommended objects; the object feature vector can be used to represent the characteristics of each object to be recommended, which may include the category, price, etc. of the object to be recommended.

[0121] like Fig. 9 As shown, a user feature vector can be constructed based on the basic information and historical behavior data of the target user, and an object feature vector can be constructed based on the object data of the object to be recommended, so as to determine the object to be recommended that matches the target user by calculating the similarity between the user feature vector and the object feature vector.

[0122] By calculating the similarity between the user feature vector and the object feature vector, the matching degree between the features of the recommended object preferred by the user and the features of the object to be recommended can be determined, so that the object to be recommended with a higher matching degree can be determined as the object to be recommended that matches the target user.

[0123] In addition, in an optional implementation, the user feature vector and the object feature vector can be generated by feature engineering. For details, refer to Fig.10 As shown, the basic information of the target user, the historical behavior data and the object data of the object to be recommended can be processed through the following steps S1010 to S1040 to generate a user feature vector and an object feature vector:

[0124] Step S1010: pre-process the basic information of the target user, historical behavior data and object data of the object to be recommended, and remove test data, dirty data and abnormal data from the above data.

[0125] Step S1020: Extract user features and object features.

[0126] Specifically, user attribute data, numerical data, and time data of the target user can be extracted. Among them, user attribute data can include user gender, address, height, identity and occupation, and interaction behavior with the object to be recommended, etc., numerical data can include age, duration of interaction behavior, etc., and time data can include date of birth, registration time, and time of interaction behavior. For the object data of the object to be recommended, object attribute data, numerical data, and time data of the object to be recommended can also be extracted. Among them, object attribute data can include object category, store, color, purpose, etc. of the object to be recommended, numerical data can include price, number of collectors, number of buyers, etc., and time data can include production date of the object to be recommended, etc.

[0127] Step S1030: Perform feature processing on the user features and object features. For example, attribute data in the user features and object features may be one-hot encoded, numerical data may be normalized and discretized, and for time data, the above features may be attenuated by determining the attenuation time.

[0128] Step S1040: Perform feature selection on the processed user features and object features to generate user feature vectors and object feature vectors. For example, the target user's gender, age, interaction behavior, time of interaction behavior, and identity and occupation can be screened out from the above user features, and the object category, purpose, object rating, price, number of purchasers, number of collections, and production date of the object to be recommended can be screened out from the above object features.

[0129] Furthermore, in order to recommend new objects to the target user, in an optional implementation, when calculating the similarity between the user feature vector and the object feature vector to determine the object to be recommended that matches the target user, similar objects to the object to be recommended can be determined based on historical behavior data, and the object to be recommended that matches the target user can be determined based on the association score between the target user and the similar objects.

[0130] Fig.11 A data processing flow chart of a second processing strategy is shown, which may include the following steps S1110 to S1160:

[0131] Step S1110. Determine the initial recommendation score of the object to be recommended based on the object data, reduce the initial recommendation score according to the decay time, and generate target feature data, which may include basic information and historical behavior data of the target user, and object data of the object to be recommended.

[0132] Step S1120. Generate a user feature vector and an object feature vector according to the above target feature data.

[0133] Step S1130: Calculate the similarity between the user feature vector and the object feature vector.

[0134] Step S1140: Use the nearest neighbor (k-Nearest Neighbor, KNN) algorithm to determine the similarity between the objects to be recommended.

[0135] Step S1150. The recommended objects whose similarity between the above-mentioned objects to be recommended is greater than a certain threshold and for which the target user has not interacted are regarded as similar objects of the objects to be recommended.

[0136] Step S1160: According to the relevance scores between the target user and the similar objects, similar objects corresponding to the first n relevance scores are determined as objects to be recommended that match the target user, where n is a positive integer.

[0137] In addition, when the features of the objects to be recommended are relatively few or are structured features, when determining the similarity between the objects to be recommended in step S1140, a classification algorithm such as a decision tree can be used to divide the similarity into multiple categories to further determine the similarity between the objects to be recommended.

[0138] Furthermore, Fig.12 The flow chart of another information processing method in this exemplary embodiment is shown as follows: Fig.12 As shown, the following steps S1201 to S1214 may be included:

[0139] Step S1201: Obtain historical behavior data of the target user and object data of the object to be recommended.

[0140] Step S1202: Determine whether the amount of historical behavior data is greater than the data amount threshold. If so, execute step S1203; if not, execute step S1209.

[0141] Step S1203: Determine that the data processing strategy for the above historical behavior data and object data is the first processing strategy.

[0142] Step S1204: Determine the initial recommendation score of the object to be recommended and the decay time of the object to be recommended based on the above historical behavior data, and then perform decay calculation on the above initial recommendation score based on the decay time to determine the current recommendation score of the object to be recommended.

[0143] Step S1205: Process the above historical behavior data according to the current recommendation score of the object to be recommended, and generate target feature data of the target user and the object to be recommended.

[0144] Step S1206: Generate a behavior feature matrix of the target user based on the above target feature data.

[0145] Step S1207: Calculate the similarity between target users based on the above behavior feature matrix, and determine similar users of the target user.

[0146] Step S1208: Determine the target user's interest in the object to be recommended based on the relevance scores between similar users and the object to be recommended.

[0147] Step S1209: Determine that the data processing strategy for the above historical behavior data and object data is the second processing strategy.

[0148] Step S1210. Determine the initial recommendation score of the object to be recommended and the decay time of the object to be recommended based on the object data, then perform decay calculation on the initial recommendation score based on the decay time to determine the current recommendation score of the object to be recommended.

[0149] Step S1211. Process the above object data according to the current recommendation score of the object to be recommended, and generate target feature data of the target user and the object to be recommended.

[0150] Step S1212: Generate a user feature vector of the target user and an object feature vector of the object to be recommended based on the above target feature data.

[0151] Step S1213: Calculate the similarity between the user feature vector and the object feature vector.

[0152] Step S1214. Determine the recommended objects that match the target user. Specifically, after determining the interest of the target user in the recommended object according to the relevance scores between similar users and the recommended object, the recommended objects corresponding to the first k interest levels are selected as the recommended objects that match the target user according to the interest of the target user in the recommended object; or the recommended objects corresponding to the first k similarities are selected as the recommended objects that match the target user according to the similarity between the user feature vector and the object feature vector.

[0153] In summary, according to the information processing method in this exemplary embodiment, the data processing strategy of the historical behavior data and the object data can be determined according to the data volume of the historical behavior data, the decay time of the object to be recommended can be determined, and the historical behavior data and the object data can be decayed according to the decay time to generate the target feature data of the target user and the object to be recommended, and then the target feature data can be processed according to the data processing strategy to determine the object to be recommended that matches the target user. On the one hand, by determining the decay time of the object to be recommended and decaying the historical behavior data and the object data according to the decay time, the dynamic change of the user's interest can be determined in combination with the influence of the time factor on the user's interest, thereby improving the accuracy of information recommendation and reducing the time it takes for users to find objects of interest; on the other hand, by determining the data processing strategy of the historical behavior data and the object data, the data processing method can be predetermined to avoid the problem of low information recommendation accuracy caused by using the same method, while also improving the efficiency of data processing.

[0154] Furthermore, this exemplary embodiment also provides an information processing device, referring to Fig.13 As shown, the information processing device 1300 may include: an acquisition module 1310, which can be used to acquire the historical behavior data of the target user and the object data of the object to be recommended; a determination module 1320, which can be used to determine the data processing strategy of the historical behavior data and the object data according to the data volume of the historical behavior data; a generation module 1330, which can be used to determine the decay time of the object to be recommended, and decay the historical behavior data and the object data according to the decay time to generate the target feature data of the target user and the object to be recommended; a processing module 1340, which can be used to process the target feature data according to the data processing strategy to determine the object to be recommended that matches the target user; wherein the decay time can be used to represent the time interval between the generation time of the initial recommendation score of the object to be recommended and the current time.

[0155] In an exemplary embodiment of the present disclosure, the determination module 1320 can be used to determine that the data processing strategy for the historical behavior data and the object data is the first processing strategy when the data volume of the historical behavior data is greater than the data volume threshold, and to determine that the data processing strategy for the historical behavior data and the object data is the second processing strategy when the data volume of the historical behavior data is not greater than the data volume threshold.

[0156] In an exemplary embodiment of the present disclosure, the generation module 1330 may be used to determine an initial recommendation score of the object to be recommended based on historical behavior data or object data, reduce the initial recommendation score according to a decay time, and generate target feature data.

[0157] In an exemplary embodiment of the present disclosure, the generation module 1330 can also be used to construct a decay function in which the current recommendation score is inversely proportional to the decay time, and reduce the initial recommendation score through the decay function to obtain the current recommendation score of the object to be recommended, and process the historical behavior data or object data according to the current recommendation score of the object to be recommended to generate target feature data.

[0158] In an exemplary embodiment of the present disclosure, the generation module 1330 may also be used to calculate the decay weight of the initial recommendation score through a decay function, and multiply the decay weight by the initial recommendation score to obtain the current recommendation score of the object to be recommended.

[0159] In an exemplary embodiment of the present disclosure, the generation module 1330 may also be used to calculate the decay weight of the initial recommendation score by the following formula:

[0160] y=ae -μΔt

[0161] Among them, △t is the decay time, a is the initial recommendation score, and μ is the decay rate.

[0162] In an exemplary embodiment of the present disclosure, the processing module 1340 can be used to process the target feature data according to a first processing strategy to obtain similar users of the target user, and determine the object to be recommended that matches the target user based on the association score between the similar users and the object to be recommended, and / or process the target feature data according to a second processing strategy to determine the object to be recommended that matches the target user.

[0163] In an exemplary embodiment of the present disclosure, the processing module 1340 can also be used to generate a behavioral feature matrix of the target user based on the target feature data, and determine similar users of the target user based on the behavioral feature matrix, determine the target user's interest in the recommended object based on the similarity between the similar users and the target user, and the association score of the similar users towards the recommended object, and determine the recommended object that matches the target user based on the interest.

[0164] In an exemplary embodiment of the present disclosure, the processing module 1340 can also be used to generate a user feature vector of the target user and an object feature vector of the object to be recommended according to the target feature data, and determine the object to be recommended that matches the target user by calculating the similarity between the user feature vector and the object feature vector.

[0165] In an exemplary embodiment of the present disclosure, the processing module 1340 may also be used to determine similar objects of the object to be recommended based on historical behavior data; and determine the object to be recommended that matches the target user based on the association score between the target user and the similar objects.

[0166] The specific details of each module in the above device have been described in detail in the implementation method of the method part. The details of the undisclosed scheme can be found in the implementation method of the method part, so they will not be repeated here.

[0167] Those skilled in the art will appreciate that various aspects of the present disclosure may be implemented as systems, methods or program products. Therefore, various aspects of the present disclosure may be specifically implemented in the following forms, namely: complete hardware implementation, complete software implementation (including firmware, microcode, etc.), or a combination of hardware and software, which may be collectively referred to herein as "circuits", "modules" or "systems".

[0168] The exemplary embodiments of the present disclosure also provide a computer-readable storage medium on which a program product capable of implementing the above-mentioned method of the present specification is stored. In some possible implementations, various aspects of the present disclosure may also be implemented in the form of a program product, which includes a program code, and when the program product is run on a terminal device, the program code is used to enable the terminal device to execute the steps according to various exemplary embodiments of the present disclosure described in the above-mentioned "Exemplary Method" section of the present specification.

[0169] The program product for implementing the above method according to the exemplary embodiment of the present disclosure may adopt a portable compact disk read-only memory (CD-ROM) and include program code, and may be run on a terminal device, such as a personal computer. However, the program product of the present disclosure is not limited thereto, and in this document, a readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, apparatus, or device.

[0170] The program product may use any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of readable storage media (a non-exhaustive list) 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.

[0171] Computer readable signal media may include data signals propagated in baseband or as part of a carrier wave, in which readable program code is carried. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Readable signal media may also be any readable medium other than a readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0172] The program code embodied on the readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination of the foregoing.

[0173] Program code for performing the operations of the present disclosure may be written in any combination of one or more programming languages, 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. In cases involving a remote computing device, the remote computing device may be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., through the Internet using an Internet service provider).

[0174] The exemplary embodiment of the present disclosure also provides an electronic device capable of implementing the above method. Fig.14 1400 according to this exemplary embodiment of the present disclosure is described. Fig.14 The electronic device 1400 shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0175] like Fig.14 As shown, the electronic device 1400 may be in the form of a general-purpose computing device. The components of the electronic device 1400 may include, but are not limited to: the at least one processing unit 1410, the at least one storage unit 1420, a bus 1430 connecting different system components (including the storage unit 1420 and the processing unit 1410), and a display unit 1440.

[0176] The storage unit 1420 stores program codes, which can be executed by the processing unit 1410, so that the processing unit 1410 executes the steps according to various exemplary embodiments of the present disclosure described in the above “Exemplary Method” section of this specification. For example, the processing unit 1410 can execute Figures 1 to 3 , Figures 5 and 6 , Figures 8 to 13 The method steps shown, etc.

[0177] The storage unit 1420 may include a readable medium in the form of a volatile storage unit, such as a random access storage unit (RAM) 1421 and / or a cache storage unit 1422 , and may further include a read-only storage unit (ROM) 1423 .

[0178] The storage unit 1420 may also include a program / utility 1424 having a set (at least one) of program modules 1425, such program modules 1425 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0179] Bus 1430 may represent one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

[0180] The electronic device 1400 may also communicate with one or more external devices 1500 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), may also communicate with one or more devices that enable a user to interact with the electronic device 1400, and / or communicate with any device that enables the electronic device 1400 to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed via an input / output (I / O) interface 1450. Furthermore, the electronic device 1400 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 1460. As shown, the network adapter 1460 communicates with other modules of the electronic device 1400 via a bus 1430. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 1400, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0181] It should be noted that, although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the exemplary embodiments of the present disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided into multiple modules or units to be embodied.

[0182] In addition, the above-mentioned figures are only schematic illustrations of the processes included in the method according to the exemplary embodiment of the present disclosure, and are not intended to be limiting. It is easy to understand that the processes shown in the above-mentioned figures do not indicate or limit the time sequence of these processes. In addition, it is also easy to understand that these processes can be performed synchronously or asynchronously, for example, in multiple modules.

[0183] Through the description of the above implementation, it is easy for those skilled in the art to understand that the exemplary implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the exemplary implementation of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method according to the exemplary implementation of the present disclosure.

[0184] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary technical means in the art that are not disclosed in the present disclosure. The specification and embodiments are to be considered exemplary only, and the true scope and spirit of the present disclosure are indicated by the claims.

Claims

1. An information processing method, characterized in that: The method comprises: Obtain the historical behavior data of the target user and the object data of the object to be recommended; Determining the data processing strategy of the historical behavior data and the object data according to the data volume of the historical behavior data; the determining the data processing strategy of the historical behavior data and the object data according to the data volume of the historical behavior data includes: when the data volume of the historical behavior data is greater than a data volume threshold, determining the data processing strategy of the historical behavior data and the object data to be a first processing strategy; when the data volume of the historical behavior data is not greater than the data volume threshold, determining the data processing strategy of the historical behavior data and the object data to be a second processing strategy; Determine the initial recommendation score of the object to be recommended according to the historical behavior data or the object data; determine the decay time of the object to be recommended according to the time interval between the generation time of the initial recommendation score of the object to be recommended and the current time; The initial recommendation score is reduced according to the decay time to generate target feature data of the target user and the object to be recommended; the target feature data is processed according to the data processing strategy to determine the object to be recommended that matches the target user; The processing of the target feature data according to the data processing strategy to determine the object to be recommended that matches the target user includes: processing the target feature data according to the first processing strategy to obtain similar users of the target user, and determining the object to be recommended that matches the target user based on the association scores between the similar users and the object to be recommended; and / or processing the target feature data according to the second processing strategy to determine the object to be recommended that matches the target user.

2. The method according to claim 1, characterized in that The step of reducing the initial recommendation score according to the decay time to generate the target feature data includes: Constructing a decay function in which the current recommendation score and the decay time are in inverse proportion, and reducing the initial recommendation score by the decay function to obtain the current recommendation score of the object to be recommended; The historical behavior data or the object data is processed according to the current recommendation score of the object to be recommended to generate the target feature data.

3. The method according to claim 2, characterized in that The step of reducing the initial recommendation score by using the attenuation function to obtain the current recommendation score of the object to be recommended includes: Calculate the decay weight of the initial recommendation score by using the decay function; The decay weight is multiplied by the initial recommendation score to obtain the current recommendation score of the object to be recommended.

4. The method according to claim 3, characterized in that The calculating the attenuation weight of the initial recommendation score by using the attenuation function includes: The decay weight of the initial recommendation score is calculated by the following formula: ； Among them, △t is the decay time, a is the initial recommendation score, and μ is the decay rate.

5. The method according to claim 1, characterized in that: The step of processing the target feature data according to the first processing strategy to obtain similar users of the target user, and determining the object to be recommended that matches the target user according to the relevance scores between the similar users and the object to be recommended, includes: Generating a behavior feature matrix of the target user according to the target feature data, and determining similar users of the target user according to the behavior feature matrix; According to the similarity between the similar users and the target user, and the relevance score of the similar users to the object to be recommended, the interest of the target user in the object to be recommended is determined, and the object to be recommended matching the target user is determined according to the interest.

6. The method according to claim 1, characterized in that The step of processing the target feature data according to the second processing strategy to determine an object to be recommended that matches the target user includes: Generating a user feature vector of the target user and an object feature vector of the object to be recommended respectively according to the target feature data; By calculating the similarity between the user feature vector and the object feature vector, an object to be recommended that matches the target user is determined.

7. The method according to claim 6, characterized in that The determining the object to be recommended that matches the target user by calculating the similarity between the user feature vector and the object feature vector includes: Determining similar objects to the object to be recommended according to the historical behavior data; An object to be recommended that matches the target user is determined according to the relevance score between the target user and the similar objects.

8. An information processing device, characterized in that: For implementing the information processing method of claim 1, the device comprises: An acquisition module is used to acquire the historical behavior data of the target user and the object data of the object to be recommended; A determination module, configured to determine a data processing strategy for the historical behavior data and the object data according to the data volume of the historical behavior data; A generating module, used to determine the decay time of the object to be recommended, and perform decay processing on the historical behavior data and the object data according to the decay time, so as to generate target feature data of the target user and the object to be recommended; The processing module is used to process the target feature data according to the data processing strategy to determine the object to be recommended that matches the target user.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

10. An electronic device, characterized in that: include: processor; as well as A memory, configured to store executable instructions of the processor; The processor is configured to perform the method of any one of claims 1 to 7 by executing the executable instructions.

Images