Information recommendation method and device

[0046] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be described clearly and completely in conjunction with the accompanying drawings in the embodiments of the present invention.

[0047] In some processes described in the specification and claims of the present invention and the above drawings, multiple operations appearing in a specific order are included, but it should be clearly understood that these operations may not be in the order in which they appear in this document. Execution or parallel execution, the operation sequence numbers such as 101, 102, etc., are only used to distinguish different operations, and the sequence numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed sequentially or in parallel. It should be noted that the descriptions of "first" and "second" in this article are used to distinguish different messages, devices, modules, etc., and do not represent a sequence, nor do they limit the "first" and "second" Are different types.

[0048] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained based on the embodiments of the present invention belong to the protection scope of the present invention.

[0049] figure 1 It is a schematic flowchart of an information recommendation method according to an embodiment of the present invention, refer to figure 1 , The method includes:

[0050] 100: According to the first consumption behavior of the users on the first platform on the second platform, the users on the first platform are divided into the first type of users and the second type of users.

[0051] In the present invention, the first platform and the second platform are different platforms, for example, the first platform is a restaurant platform, and the second platform is a supermarket platform; for another example, the first platform is an e-commerce platform such as Taobao, The second platform is a catering platform such as takeaway.

[0052] In the present invention, compared to the second platform, the first platform has more complete information in at least one dimension, and the dimensions include user group, number of users, user information, user behavior records, and the like.

[0053] 102: According to the user characteristics of the first type of users and the second type of users, select similar users of the first type of users from the second type of users.

[0054] Optionally, in an implementation manner of this embodiment, the user characteristics of the first type of user and the second type of user may be predetermined or set.

[0055] Optionally, in an implementation manner of this embodiment, the user characteristics of the first type of user and the second type of user refer to the fact that the first type of user and the second type of user are in the first type of user User characteristics on the platform, that is, user characteristics extracted based on relevant information on the first platform.

[0056] 104: Determine a recommendation object of the second type of user on the second platform based on the second consumption behavior of the similar user on the second platform. In other words, the objects on the second platform are recommended to the second type of users.

[0057] Using the information recommendation method provided by this embodiment, based on the similarity between users on the first platform and the consumption behaviors of some users on the first platform, recommend objects on the second platform for other users on the first platform (For example, coupons for goods, services, goods and services, etc.), it has the following effects: recommendation based on similar users, so that even cross-platform, the information recommendation results are also targeted; by calculating the second category of users of each user Similar users makes the number of users covered by the final recommendation flexible and controllable. If necessary, it can cover all the second-type users; even in scenarios where there is less user data on the second platform (the so-called cold start), Recommendations can be made for the second type of users, which can solve the cold start problem caused by fewer initial users of existing platforms.

[0058] Optionally, in an implementation manner of this embodiment, in the process 100, the first consumption behavior includes: whether to place an order and/or the number of orders. In other words, in the process 100, according to whether a user on the first platform places an order on the second platform, or whether the number of orders placed by a user on the first platform reaches a set value on the second platform, the Users are divided into the first type of users and the second type of users. Wherein, the first type of users are users who have placed orders on the second platform or placed orders exceeding a set number, and the second type of users are users who have not placed orders on the second platform or placed orders that have not exceeded the set number.

[0059] Optionally, in an implementation manner of this embodiment, in the processing 104, the second consumption behavior includes the order object, or the order object and the order quantity. Specifically, in processing 104, the recommendation objects of the second type of users (for example, goods, services, coupons for goods and services, etc.) can be determined according to the order objects of similar users of the second type of users on the second platform. ; It is also possible to synthesize the order objects and the number of orders placed by similar users of the second type of users on the second platform to determine the recommended objects of the second type of users.

[0060] In the above two implementation manners, the first consumption behavior is used to classify users, and the second consumption behavior is used to determine and screen recommended objects. On this basis, combined with the contents disclosed in the embodiments of the present invention and its implementation, those skilled in the art should understand that the first consumption behavior and the second consumption behavior can be set flexibly and reasonably to achieve the above objectives, which also falls on Within the protection scope of the present invention.

[0061] Optionally, in an implementation manner of this embodiment, as figure 2 As shown, processing 102 is implemented in the following manner:

[0062] 1020: Determine the user characteristics of the first type of user and the second type of user on the first platform according to the consumption behavior of the first type of user and the second type of user on the first platform .

[0063] 1022: Calculate the similarity between each user in the first type user and each user in the second type user based on the user characteristics of the first type user and the second type user on the first platform.

[0064] 1024: Determine similar users of each user in the second type of users based on the similarity.

[0065] With this implementation, similar users of the second type of users can be calculated based on the user characteristics on the first platform. Since the user characteristics are derived from the same platform, the similarity can more accurately reflect the similarity between users, which helps to improve the accuracy of the recommendation results in a cross-platform recommendation scenario.

[0066] Optionally, in an implementation manner of this embodiment, a second type of user may have multiple first type users with the same similarity. In this case, in processing 104, these first type users with the same similarity may be The order objects of the users of the second category are also recommended for the users of the second category; the order objects of the first category users with the same similarity can also be compared to these order objects according to their current preferential strength, popularity, and praise. Sort, and then select one or more as the recommended objects of the second type of user according to the sorting result.

[0067] image 3 It is a schematic flowchart of an information recommendation method according to an embodiment of the present invention. Reference image 3 , The method includes:

[0068] 300: According to the first consumption behavior of the users on the first platform on the second platform, the users on the first platform are divided into the first type of users and the second type of users.

[0069] In the present invention, the first platform and the second platform are different platforms, for example, the first platform is a restaurant platform, and the second platform is a supermarket platform; for another example, the first platform is an e-commerce platform such as Taobao, The second platform is a catering platform such as takeaway.

[0070] In the present invention, compared to the second platform, the first platform has more complete information in at least one dimension, and the dimensions include user group, number of users, user information, user behavior records, and the like.

[0071] 302: Select similar users of the first type of users from the second type of users according to the user characteristics of the first type of users and the second type of users.

[0072] Optionally, in an implementation manner of this embodiment, the user characteristics of the first type of user and the second type of user may be predetermined or set.

[0073] Optionally, in an implementation manner of this embodiment, the user characteristics of the first type of user and the second type of user refer to the fact that the first type of user and the second type of user are in the first type of user User characteristics on the platform, that is, user characteristics extracted based on relevant information on the first platform.

[0074] 304: Determine a recommendation object of the second type of user on the second platform based on the second consumption behavior of the similar user on the second platform. In other words, the objects on the second platform are recommended to the second type of users.

[0075] 306: Update the first type of user and the second type of user according to the first consumption behavior of the second type of user on the second platform.

[0076] Optionally, in an implementation of this embodiment, in processing 306, if there are users of the second type who place orders on the second platform, or there are users whose number of orders on the second platform meets the requirements, The user is removed from the second category of users and added to the first category of users.

[0077] Using the information recommendation method provided in this embodiment, except for having figure 1 In addition to the effects of the illustrated embodiment and its implementation, the pertinence of subsequent recommendations can be improved by updating the first-type users and the second-type users. In addition, it is beneficial to further analyze user behavior based on changes in the first type of user or the second type of user.

[0078] Optionally, in an implementation manner of this embodiment, the processing 300-306 may be periodically executed in a loop. For example, processing 300-306 is performed for each recommendation.

[0079] In this embodiment, for the detailed description of the processing 300-304, reference may be made to the description of the processing 100-104 above, which will not be repeated here.

[0080] Figure 4 A schematic flowchart of an information recommendation method according to an embodiment of the present invention is shown. Reference Figure 4 Taking the first platform as the catering platform and the second platform as the supermarket platform as an example, the method includes:

[0081] 400: User screening, locating seed users.

[0082] Wherein, the seed user is a user who has placed an order on the supermarket platform within three months (customization) among the catering platform users.

[0083] 402: Extract user characteristics of catering platform users.

[0084] Optionally, in an implementation of this embodiment, the historical order characteristics of all users are extracted, including: the number of historical orders, the number of stores, the number of visits, the frequency of opening applications, the number of group meals, and price sensitivity Features such as degree, order rate, etc.; extract the attributes of users themselves, including gender, age, occupation, income, education and other characteristics.

[0085] 404: Calculate the similarity between users based on the law of cosines.

[0086] Divide all users into two types, one is a supermarket user, which is also a seed user, and the other is a non-commercial supermarket user, that is, the catering platform users have not been downloaded on the supermarket platform within three months (custom) Single user.

[0087] According to the feature vector constructed based on the user characteristics, calculate the cosine distance between the supermarket user and the non-commercial supermarket user. For each non-commercial supermarket user, select the supermarket user with the highest similarity as the non-commercial supermarket user Of similar users.

[0088] 406: Information recommendation.

[0089] Optionally, in an implementation of this embodiment, according to the order data (including: the store where the order was placed, the product ordered, the number of orders, etc.) of similar users who are non-commercial super users on the supermarket platform, Recommend objects for non-supermarket users (for example, commodities, services, coupons for commodities and services, etc.).

[0090] Optionally, in an implementation of this embodiment, when calculating the similarity between a non-commercial supermarket user and a commercial supermarket user, the supermarket merchants that the supermarket user has placed an order are recorded behind the non-commercial supermarket user, and The similarity between the two users is used as the rating of this business. Therefore, taking the issuance of coupons as an example, after data analysis, there will be one or more merchants behind each user, and each merchant will also have a similarity score. Sort according to this similarity score, according to the score The size is used as the priority to select one or more merchants' coupons and issue them to the corresponding users.

[0091] The method provided in this embodiment can not only provide users with targeted recommendation objects with flexible user coverage, but also solve the cold start problem of the supermarket platform in the early stage of construction due to the small number of users.

[0092] Figure 5 It is a block diagram of an information recommendation device according to an embodiment of the present invention. Reference Figure 5 The information recommendation device includes a classification module 50, a similarity determination module 52, and an object determination module 54. The details are described below.

[0093] In this embodiment, the classification module 50 is configured to classify users on the first platform into a first type of user and a second type of user according to the first consumption behavior of the user on the first platform on the second platform.

[0094] In the present invention, the first platform and the second platform are different platforms, for example, the first platform is a restaurant platform, and the second platform is a supermarket platform; for another example, the first platform is an e-commerce platform such as Taobao, The second platform is a catering platform such as takeaway.

[0095] In the present invention, compared to the second platform, the first platform has more complete information in at least one dimension, and the dimensions include user group, number of users, user information, user behavior records, and the like.

[0096] In this embodiment, the similarity determination module 52 is configured to select similar users of the first type of users from the second type of users according to the user characteristics of the first type of users and the second type of users.

[0097] Optionally, in an implementation manner of this embodiment, the user characteristics of the first type of user and the second type of user may be predetermined or set.

[0098] Optionally, in an implementation manner of this embodiment, the user characteristics of the first type of user and the second type of user refer to the fact that the first type of user and the second type of user are in the first type of user User characteristics on the platform, that is, user characteristics extracted based on relevant information on the first platform.

[0099] In this embodiment, the object determination module 54 is configured to determine the recommended object of the second type of user on the second platform based on the second consumption behavior of the similar users on the second platform.

[0100] Using the information recommendation device provided in this embodiment, based on the similarity between users on the first platform and the consumption behavior of some users on the first platform, recommend objects on the second platform for other users on the first platform (For example, coupons for goods, services, goods and services, etc.), it has the following effects: recommendation based on similar users, so that even cross-platform, the information recommendation results are also targeted; by calculating the second category of users of each user Similar users makes the number of users covered by the final recommendation flexible and controllable. If necessary, it can cover all the second-type users; even in scenarios where there is less user data on the second platform (the so-called cold start), Recommendations can be made for the second type of users, which can solve the cold start problem caused by fewer initial users of existing platforms.

[0101] Optionally, in an implementation manner of this embodiment, the first consumption behavior includes: whether to place an order and/or the number of orders. In other words, the classification module 50 can classify users on the first platform according to whether users on the first platform place orders on the second platform, or according to whether the number of orders placed by users on the first platform on the second platform reaches a set value. For the first type of user and the second type of user. Wherein, the first type of users are users who have placed orders on the second platform or placed orders exceeding a set number, and the second type of users are users who have not placed orders on the second platform or placed orders that have not exceeded the set number.

[0102] Optionally, in an implementation manner of this embodiment, the second consumption behavior includes the order object, or the order object and the order quantity. Specifically, the object determination module 54 may determine the recommended objects of the second type of users (for example, goods, services, coupons for goods and services, etc.) according to the order objects of similar users of the second type of users on the second platform; It is also possible to integrate the order objects and the number of orders placed by similar users of the second type of users on the second platform to determine the recommended objects of the second type of users.

[0103] In the above two implementation manners, the first consumption behavior is used to classify users, and the second consumption behavior is used to determine and screen recommended objects. On this basis, combined with the contents disclosed in the embodiments of the present invention and its implementation, those skilled in the art should understand that the first consumption behavior and the second consumption behavior can be set flexibly and reasonably to achieve the above objectives, which also falls on Within the protection scope of the present invention.

[0104] Optionally, in an implementation manner of this embodiment, as Image 6 As shown, the similarity determination module 54 includes:

[0105] The feature submodule 540 is configured to determine whether the first type of user and the second type of user are in the first type of user according to the consumption behavior of the first type of user and the second type of user on the first platform. User characteristics on the platform.

[0106] The similarity sub-module 542 is configured to calculate each user in the first type of user and the second type of user based on the user characteristics of the first type of user and the second type of user on the first platform The similarity of each user.

[0107] The similarity confirmation sub-module 544 is configured to determine similar users of each user in the second type of users based on the similarity.

[0108] With this implementation, similar users of the second type of users can be calculated based on the user characteristics on the first platform. Since the user characteristics are derived from the same platform, the similarity can more accurately reflect the similarity between users, which helps to improve the accuracy of the recommendation results in a cross-platform recommendation scenario.

[0109] Figure 7 It is a block diagram of an information recommendation device according to an embodiment of the present invention. Reference Figure 7 The information recommendation device includes a classification module 50, a similarity determination module 52, an object determination module 54 and an update module 56.

[0110] In this embodiment, for the detailed description of the classification module 50, the similarity determination module 52, and the object determination module 54, please refer to the foregoing, which will not be repeated here.

[0111] In this embodiment, the update module 56 is configured to update the first type of user and the second type of user according to the first consumption behavior of the second type of user on the second platform.

[0112] Optionally, in an implementation manner of this embodiment, if there are users who place orders on the second platform among the second type of users, or there are users whose number of orders on the second platform meets the requirements, the update module 56 will The user is removed from the second category of users and added to the first category of users.

[0113] Using the information recommendation device provided in this embodiment, except for having Figure 5 In addition to the effects of the illustrated embodiment and its implementation, the pertinence of subsequent recommendations can be improved by updating the first-type users and the second-type users. In addition, it is beneficial to further analyze user behavior based on changes in the first type of user or the second type of user.

[0114] Optionally, in an implementation manner of this embodiment, the information recommendation apparatus calls the classification module 50, the similarity determination module 52, the object determination module 54, and the update module 56 cyclically.

[0115] In the present invention, in addition to the above information recommendation method and information recommendation device, an embodiment of the present invention also provides a computer storage medium, which stores one or more computer instructions, wherein the one or more When computer instructions are executed Figure 1-Figure 4 A method provided by any one of the illustrated embodiments or implementations.

[0116] Figure 8 It is a block diagram of an electronic device according to an embodiment of the present invention. Reference Figure 8 , The electronic device includes a memory 80 and a processor 82. The number of the memory 80 and the processor 82 may be one or more. Wherein, the one or more memories 80 store one or more computer instructions for the one or more processors 82 to call and execute; the one or more processors are used to execute the one or more When multiple computer instructions are implemented, such as Figure 1-Figure 4 A method provided by any one of the illustrated embodiments or implementations.

[0117] Optionally, in an implementation manner of this embodiment, as Figure 8 As shown by the dashed box, the electronic device also includes an input and output interface for data communication.

[0118] Those skilled in the art can clearly understand that, for the convenience and conciseness of the description, the specific working process of the above-described system, device, and unit can refer to the corresponding process in the foregoing method embodiment, which will not be repeated here.

[0119] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in One place, or it can be distributed to multiple network units. Some or all of the modules may be selected according to actual needs to achieve the objectives of the solutions of the embodiments. Those of ordinary skill in the art can understand and implement without creative work.

[0120] Through the description of the above implementation manners, those skilled in the art can clearly understand that each implementation manner can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution essentially or the part that contributes to the existing technology can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM/RAM, magnetic A disc, an optical disc, etc., include a number of instructions to make a computer device (which may be a personal computer, a server, or a network device, etc.) execute the methods described in each embodiment or some parts of the embodiment.

[0121] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: The technical solutions recorded in the foregoing embodiments are modified, or some of the technical features are equivalently replaced; these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

[0122] The present invention discloses A1, an information recommendation method, including:

[0123] Dividing users on the first platform into a first type of user and a second type of user according to the first consumption behavior of the user on the first platform on the second platform;

[0124] Selecting similar users of the first type of users from the second type of users according to the user characteristics of the first type of users and the second type of users;

[0125] Determine a recommendation object of the second type of user on the second platform based on the second consumption behavior of the similar users on the second platform.

[0126] A2, according to the method described in A1,

[0127] The first consumption behavior includes: whether to place an order and/or the number of orders;

[0128] The second consumption behavior includes the order object, or the order object and the order quantity.

[0129] A3. The method according to A1, which includes:

[0130] The first type of user and the second type of user are updated according to the first consumption behavior of the second type of user on the second platform.

[0131] A4. The method according to any one of A1-A3, wherein the first-type user is selected from the second-type user according to the user characteristics of the first-type user and the second-type user Of similar users, including:

[0132] Determining the user characteristics of the first type of user and the second type of user on the first platform according to the consumption behavior of the first type of user and the second type of user on the first platform;

[0133] Calculating the similarity between each user in the first type user and each user in the second type user based on the user characteristics of the first type user and the second type user on the first platform;

[0134] Determine similar users of each user in the second type of users based on the similarity.

[0135] The present invention also discloses B5, an information recommendation device, including:

[0136] The classification module is configured to classify users on the first platform into a first type of user and a second type of user according to the first consumption behavior of the user on the first platform on the second platform;

[0137] A similarity determination module, configured to select similar users of the first type of users from the second type of users according to the user characteristics of the first type of users and the second type of users;

[0138] The object determination module is configured to determine the recommended object of the second type of user on the second platform based on the second consumption behavior of the similar users on the second platform.

[0139] B6. The device as described in B5,

[0140] The first consumption behavior includes: whether to place an order and/or the number of orders;

[0141] The second consumption behavior includes the order object, or the order object and the order quantity.

[0142] B7. The device as described in B5, further comprising:

[0143] The update module is configured to update the first type of user and the second type of user according to the first consumption behavior of the second type of user on the second platform.

[0144] B8. The device according to any one of B5-B7, wherein the similarity determination module includes:

[0145] The feature submodule is configured to determine whether the first type of user and the second type of user are on the first platform according to the consumption behavior of the first type of user and the second type of user on the first platform User characteristics on

[0146] The similarity sub-module is used to calculate each user in the first type of user and each of the second type of users based on the user characteristics of the first type of users and the second type of users on the first platform User similarity;

[0147] The similarity confirmation sub-module is configured to determine similar users of each user in the second type of users based on the similarity.

[0148] The present invention also discloses C9, a computer storage medium that stores one or more computer instructions, where the one or more computer instructions implement the method described in any one of A1-A4 when executed.

[0149] The present invention also discloses D10, an electronic device including one or more memories and one or more processors,

[0150] Wherein, the one or more memories stores one or more computer instructions for the one or more processors to call and execute;

[0151] Wherein, the one or more processors are used to implement the method as described in any one of A1-A4 when executing the one or more computer instructions.