Information recommendation method and device, and computer storable medium

Abstract

The present disclosure relates to an information recommendation method and device, a computer storable medium, and the field of computer technology. The information recommendation method includes: determining at least one target keyword and its target vector based on the keyword to be processed, with the target keyword corresponding to the target vector in one-to-one correspondence; for each target keyword, obtaining multiple reference keywords belonging to the same keyword category as each target keyword from historical data corresponding to the object related to the keyword to be processed; determining a comprehensive vector corresponding to the object based on multiple reference vectors corresponding to the multiple reference keywords, with the reference keywords corresponding to the reference vectors in one-to-one correspondence; and recommending information related to the keyword to be processed based on the similarity between the comprehensive vector and each target vector. According to the present disclosure, the accuracy of recommendations can be improved.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and in particular to an information recommendation method and device, and a computer storable medium. Background Art

[0002] With the rapid development of the live streaming industry, how to accurately screen users to sell goods through live streaming has become the key to promoting the development of the live streaming industry.

[0003] In the related technology, for an item to be promoted, the promotion frequency and sales volume of different promotion users are counted, and the merchants are recommended promotion users with high promotion frequency and high sales volume. Summary of the Invention

[0004] In the related art, the matching degree between the user who brings goods and the items to be brought is not taken into consideration, so the accuracy of the recommendation is poor.

[0005] In response to the above technical problems, the present disclosure proposes a solution that can improve the accuracy of recommendations.

[0006] According to a first aspect of the present disclosure, an information recommendation method is provided, comprising: determining at least one target keyword and its target vector based on a keyword to be processed, wherein the target keyword and the target vector have a one-to-one correspondence; for each target keyword, obtaining a plurality of reference keywords belonging to the same keyword category as each target keyword from historical data corresponding to an object related to the keyword to be processed; determining a comprehensive vector corresponding to the object based on a plurality of reference vectors corresponding to the plurality of reference keywords, wherein the reference keyword and the reference vector have a one-to-one correspondence; and recommending information related to the keyword to be processed based on a similarity between the comprehensive vector and each target vector.

[0007] In some embodiments, determining the comprehensive vector corresponding to the object includes: for the object, determining conversion data corresponding to each reference keyword based on corresponding historical data; and determining the comprehensive vector based on the multiple reference vectors and their corresponding conversion data.

[0008] In some embodiments, determining the comprehensive vector based on the multiple reference vectors and their corresponding conversion data includes: determining a corresponding conversion weight based on the conversion data corresponding to each reference keyword; and performing a weighted operation on the multiple reference vectors using the multiple conversion weights to obtain the comprehensive vector.

[0009] In some embodiments, the object is a user who sells goods, and the historical data is historical sales data. The historical sales data of the user who sells goods include the sales data of each day in a historical time period. The daily sales data include at least one of the number of clicks, order volume, number of click-through fans, and number of online fans of the sold items corresponding to each reference keyword in one day. The number of click-through fans is the number of fans who click on the sold items among the online fans of the user who sells goods in one day. The conversion data includes at least one of the click-through rate, conversion rate, and fan activity rate of each day in the historical time period. The click-through rate is positively correlated with the number of clicks and negatively correlated with the number of online fans. The conversion rate is positively correlated with the number of orders and negatively correlated with the number of clicks. The fan activity rate is the ratio between the number of click-through fans and the number of online fans.

[0010] In some embodiments, there are multiple users who bring goods to the market, and the click-through rate is positively correlated with the average number of clicks and negatively correlated with the average number of online fans. The average number of clicks is the average of the number of clicks of the multiple users who bring goods to the market, and the average number of online fans is the average of the number of online fans of the multiple users who bring goods to the market; the conversion rate is positively correlated with the average order volume and negatively correlated with the average click volume. The average order volume is the average of the order volumes of the multiple users who bring goods to the market.

[0011] In some embodiments, determining the corresponding conversion weight includes: determining an average of the click-through rate, the conversion rate, and the fan activity rate as the conversion weight.

[0012] In some embodiments, the historical data corresponding to the object includes at least one set of order data derived from the object, each set of order data includes item identifiers of multiple items, and obtaining multiple reference keywords belonging to the same keyword category as each target keyword from the historical data corresponding to the object related to the keyword to be processed includes: determining multiple item groups based on each set of order data derived from the object, each item group including item identifiers of two different items; determining the association weight value between two items corresponding to the item identifier in each item group based on each set of order data derived from the object, the larger the association weight value, the stronger the correlation between the two items; constructing a graph model with the item identifiers in the multiple item groups as vertices and the association weight values ​​as edges; and using the graph model, obtaining multiple reference keywords belonging to the same keyword category as each target keyword from the historical data corresponding to the object.

[0013] In some embodiments, the order data also includes other order parameters corresponding to the item identifier, and using the graph model to obtain multiple reference keywords that belong to the same keyword category as each target keyword from the historical data corresponding to the object includes: using a random walk algorithm to perform a random walk on the graph model until the walk stop condition is met, and obtaining at least one item identifier as a reference item identifier; for each target keyword, from each group of order data derived from the object, selecting an item identifier or other order parameter that belongs to the same keyword category as each target keyword and corresponds to the reference item identifier as a reference keyword.

[0014] In some embodiments, the order data also includes an order date, and determining the association weight value between two items corresponding to the item identifiers in each item group includes: determining the association weight value between the two items corresponding to the item identifiers in each item group based on the difference between the order dates corresponding to the two item identifiers in each item group, and the association weight value is negatively correlated with the difference between the order dates.

[0015] In some embodiments, the order data also includes an order date. Determining multiple groups of items based on the order data corresponding to each user who brings goods includes: determining at least one item set based on each group of order data, each item set including multiple item identifiers arranged in order of order, and the difference between the latest order date and the earliest order date corresponding to the multiple item identifiers in the same item set is less than a date threshold; for each item set, dividing two adjacent item identifiers into one item group to obtain the multiple item groups.

[0016] In some embodiments, recommending information related to the keyword to be processed includes: when the object is a plurality of objects, selecting an object to be recommended from the plurality of objects, and recommending relevant information of the object to be recommended.

[0017] In some embodiments, when the object is one object, the at least one target keyword includes multiple target keywords, and recommending information related to the keyword to be processed includes: when the object is one object, selecting the keyword to be recommended from the multiple target keywords, and recommending relevant information of the keyword to be recommended.

[0018] In some embodiments, determining at least one target keyword based on the keyword to be processed includes: converting the keyword to be processed into at least one target keyword based on a keyword category corresponding to the keyword to be processed.

[0019] In some embodiments, when the object is a user who brings goods, the keyword categories include user identification category, item identification category, brand identification category, store identification category and category.

[0020] According to a second aspect of the present disclosure, an information recommendation device is provided, comprising: a first determination module, configured to determine at least one target keyword and its target vector based on the keyword to be processed, wherein the target keyword and the target vector have a one-to-one correspondence; an acquisition module, configured to obtain, for each target keyword, a plurality of reference keywords belonging to the same keyword category as each target keyword from historical data corresponding to an object related to the keyword to be processed; a second determination module, configured to determine a comprehensive vector corresponding to the object based on a plurality of reference vectors corresponding to the plurality of reference keywords, wherein the reference keywords and the reference vectors have a one-to-one correspondence; and a recommendation module, configured to recommend information related to the keyword to be processed based on a similarity between the comprehensive vector and each target vector.

[0021] According to a third aspect of the present disclosure, an information recommendation device is provided, comprising: a memory; and a processor coupled to the memory, wherein the processor is configured to execute the information recommendation method described in any of the above embodiments based on instructions stored in the memory.

[0022] According to a fourth aspect of the present disclosure, a computer storable medium is provided, on which computer program instructions are stored. When the instructions are executed by a processor, the information recommendation method described in any of the above embodiments is implemented.

[0023] In the above embodiment, the accuracy of recommendation can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate embodiments of the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0025] The present disclosure can be more clearly understood from the following detailed description with reference to the accompanying drawings, in which:

[0026] Figure 1 is a flowchart illustrating an information recommendation method according to some embodiments of the present disclosure;

[0027] Figure 2 is a flowchart illustrating obtaining a plurality of reference keywords belonging to the same keyword category as each target keyword according to some embodiments of the present disclosure;

[0028] Figure 3 is a flow chart illustrating determining a comprehensive vector corresponding to an object according to some embodiments of the present disclosure;

[0029] Figure 4is a block diagram illustrating an information recommendation apparatus according to some embodiments of the present disclosure;

[0030] Figure 5 is a block diagram illustrating an information recommendation device according to some other embodiments of the present disclosure;

[0031] Figure 6 is a block diagram illustrating a computer system for implementing some embodiments of the present disclosure. DETAILED DESCRIPTION

[0032] Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. It should be noted that unless otherwise specifically stated, the relative arrangement of components and steps, numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present disclosure.

[0033] At the same time, it should be understood that for the convenience of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship.

[0034] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the present disclosure, its application, or uses.

[0035] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered part of the specification.

[0036] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0037] It should be noted that like reference numerals and letters refer to like items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0038] Figure 1 is a flowchart illustrating an information recommendation method according to some embodiments of the present disclosure.

[0039] like Figure 1As shown, the information recommendation method includes: step S10, determining at least one target keyword and its target vector based on the keyword to be processed; step S20, for each target keyword, obtaining multiple reference keywords belonging to the same keyword category as each target keyword from the historical data corresponding to the object related to the keyword to be processed; step S30, determining the comprehensive vector corresponding to the object based on the multiple reference vectors corresponding to the multiple reference keywords; step S40, recommending information related to the keyword to be processed based on the similarity between the comprehensive vector and each target vector. In some embodiments, the information recommendation method is executed by an information recommendation device. The information recommendation method can be applied to live streaming scenarios, as well as to advertising media selection scenarios.

[0040] In step S10, at least one target keyword and its target vector are determined based on the keyword to be processed. The target keyword and the target vector correspond one to one. In some embodiments, the keyword to be processed can be a search term input by the search user.

[0041] In some embodiments, for example, the keywords to be processed include, but are not limited to, the item identifier of the brought item, the user identifier of the user who brought the item, the brand identifier of the brought item, and the category identifier of the brought item. For example, the item identifier can be an item ID (Identity Document) or the item name, the user identifier can be a user ID or a user name, the brand identifier can be a brand ID or a brand name, and the category identifier can be a category ID or a category name.

[0042] In some embodiments, the keyword to be processed is converted into at least one target keyword based on the keyword category corresponding to the keyword to be processed. In some embodiments, when the subject associated with the keyword to be processed is a user who brings in goods, the keyword categories include but are not limited to user identification, item identification, brand identification, store identification, and category.

[0043] For example, when the keyword category corresponding to the keyword to be processed is the item identification category, the keyword to be processed itself is used as the target keyword.

[0044] For another example, if the keyword category corresponding to the keyword to be processed is a user identification category, the keyword to be processed is converted into at least one of an item identification, a category identification, and a brand identification corresponding to the user identification corresponding to the keyword to be processed. In some embodiments, a single user identification corresponds to multiple item identifications, multiple category identifications, or multiple brand identifications. In this case, the target vector corresponding to the target keyword composed of the multiple item identifications can be the average or weighted average of the multiple identification vectors of the multiple item identifications. Similarly, the target vector corresponding to the target keyword composed of multiple category identifications or multiple brand identifications can be determined.

[0045] For another example, when the keyword category corresponding to the keyword to be processed is a user identification category, the keyword to be processed may also be converted into item identifications of a plurality of candidate items.

[0046] In step S20 , for each target keyword, a plurality of reference keywords belonging to the same keyword category as each target keyword are obtained from historical data corresponding to objects related to the keyword to be processed.

[0047] For example, by Figure 2 The above step S20 is implemented in the manner shown.

[0048] Figure 2 is a flowchart illustrating obtaining a plurality of reference keywords belonging to the same keyword category as each target keyword according to some embodiments of the present disclosure.

[0049] like Figure 2 As shown, obtaining multiple reference keywords belonging to the same keyword category as each target keyword includes steps S21 to S24. For example, the historical data corresponding to the above-mentioned object includes at least one set of order data derived from the object. Each set of order data includes item identifiers of multiple items. For example, in the case where the object is a user who sells goods, each set of order data derived from the object is the order data of each fan of the user who sells goods. The multiple items included in each set of order data are multiple sold goods.

[0050] For example, the order data of fan A can be represented as {[Product ID=3645, Date=20200101, Brand ID=Jordan, Store=Jordan Store, Category=Sports Shoes], [Product ID=4543, Date=20200101, Brand ID=Nike, Store=Nike Store, Category=Sports Shoes], [Product ID=712345, Date=20200102, Brand ID=Huawei, Store=Huawei Store, Category=Mobile Phones], [Product ID=546574, Date=20200108, Brand ID=Picardin, Store=Picardin Store, Category=Jeans]}.

[0051] In step S21, multiple item groups are determined based on each set of order data derived from the object. Each item group includes the item identifiers of two different items. For example, if the order data includes order dates, at least one item set is determined based on each set of order data. Each item set includes multiple item identifiers arranged in order order, and the difference between the latest order date and the earliest order date corresponding to the multiple item identifiers in the same item set is less than a date threshold (e.g., 3 days). For each item set, two adjacent item identifiers are grouped together to form multiple item groups.

[0052] For example, fan A's order data can be divided into order set B and order set C. Order set B is {[Product ID = 3645, Date = 20200101, Brand ID = Jordan, Store = Jordan Store, Category = Sports Shoes], [Product ID = 4543, Date = 20200101, Brand ID = Nike, Store = Nike Store, Category = Sports Shoes], [Product ID = 712345, Date = 20200102, Brand ID = Huawei, Store = Huawei Store, Category = Mobile Phones]}. Order set C is {[Product ID = 546574, Date = 20200108, Brand ID = Picardin, Store = Picardin Store, Category = Jeans], [Product ID = 546578, Date = 20200109, Brand ID = Picardin, Store = Picardin Store, Category = Denim Jackets]}. The difference between the latest order date and the earliest order date in a single order set is less than 3 days. Based on these two order sets, two item sets can be determined: {item ID = 3645, item ID = 4543, item ID = 712345} and {item ID = 546574, item ID = 546578}.

[0053] In step S22, based on each set of order data derived from the object, an association weight value between two items corresponding to the item identifiers in each item group is determined. The larger the association weight value, the stronger the correlation between the two items.

[0054] In some embodiments, the order data further includes order dates. An association weight value between two items corresponding to the item identifiers in each item group is determined based on the difference between the order dates corresponding to the two item identifiers in each item group. The association weight value is negatively correlated with the difference between the order dates.

[0055] For example, the association weight value is D is the difference between the order dates (the later order date minus the earlier order date), and M is a preset positive integer. For example, M is set to 4. This is because the difference between the order dates will not exceed 4.

[0056] In step S23, a graph model is constructed using the item identifiers in multiple item groups as vertices and the associated weight values ​​as edges. For example, after obtaining the associated weight values, order set B is processed to obtain the following processed sets: {[Product ID = 3645, Date = 20200101, Brand ID = Jordan, Store = Jordan Store, Category = Sports Shoes], [Product ID = 4543, Date = 20200101, Brand ID = Nike, Store = Nike Store, Category = Sports Shoes], Weight = 1}, and {[Product ID = 4543, Date = 20200101, Brand ID = Nike, Store = Nike Store, Category = Sports Shoes], [Product ID = 712345, Date = 20200102, Brand ID = Huawei, Store = Huawei Store, Category = Mobile Phones], Weight = 0.75}. A graph model can be constructed based on these processed sets.

[0057] In step S24 , a plurality of reference keywords belonging to the same keyword category as each target keyword are acquired from the historical data corresponding to the object using the graph model.

[0058] In some embodiments, the order data also includes other order parameters corresponding to the item identifier. The above step S24 can be implemented as follows.

[0059] First, a random walk algorithm is used to perform a random walk on the graph model until a stopping condition is met, obtaining at least one item identifier as a reference item identifier. For example, the parameters of the random walk include the maximum step length (e.g., 13) and the number of traversals of each vertex (e.g., 4). The stopping condition is the total length of the walk. In some embodiments, the random walk algorithm is a random walk algorithm based on alias sampling.

[0060] Then, for each target keyword, from each set of order data derived from the object, an item ID or other order parameter that belongs to the same keyword category as each target keyword and corresponds to a reference item ID is selected as the reference keyword. Taking the order data including item ID, brand ID, and category ID as an example, if the target keyword belongs to the item ID category, the reference keyword is also the item ID (i.e., the reference item ID). If the target keyword belongs to the brand ID category, the reference keyword is the brand ID corresponding to the reference item ID. If the target keyword belongs to the category ID category, the reference keyword is the category ID corresponding to the reference item ID.

[0061] In the above embodiment, by calculating the association weight value and constructing the graph model, the association between items can be constructed, and then the association between items can be used to determine the reference keyword most associated with the target keyword, thereby further improving the accuracy of the recommendation.

[0062] return Figure 1 In step S30, a comprehensive vector corresponding to the object is determined based on multiple reference vectors corresponding to multiple reference keywords. Taking the object as a user who brings goods as an example, each user who brings goods corresponds to a comprehensive vector.

[0063] For example, by Figure 3 The above step S30 is implemented in the manner shown.

[0064] Figure 3 is a flow chart illustrating determining a synthetic vector corresponding to an object according to some embodiments of the present disclosure.

[0065] like Figure 3 As shown, determining the comprehensive vector corresponding to the object includes steps S31 and S32.

[0066] In step S31, the conversion data corresponding to each reference keyword is determined based on the historical data corresponding to the object. The conversion data is data that measures the ability to generate specific contributions (for example, traffic contributions such as click-through rate, conversion rate, and fan activity rate) based on each reference keyword. In some embodiments, when the object is a user who brings goods, the historical data is historical goods-bringing data, and the conversion data is goods-bringing conversion data. In other embodiments, when the object is an advertising media, the historical data is historical advertising data, and the conversion data is advertising conversion data.

[0067] In step S32, a comprehensive vector is determined based on the multiple reference vectors and their corresponding conversion data. By combining the conversion data to determine the comprehensive vector, not only the matching degree between the keywords to be processed and the object is considered, but also the conversion ability of the object is considered, which can further improve the accuracy of the recommendation.

[0068] In some embodiments, a corresponding conversion weight is determined based on the conversion data corresponding to each reference keyword. Using multiple conversion weights, a weighted operation is performed on multiple reference vectors to generate a composite vector. By converting conversion data into conversion weights and performing a weighted operation on multiple reference vectors, a comprehensive evaluation of the object's conversion ability can be achieved, further improving the accuracy of recommendations. The conversion weight is the weight value of each reference vector.

[0069] For example, if the user is an influencer and the historical data is historical influencer sales data, the influencer's historical influencer sales data includes daily sales data within a historical period. Daily sales data includes at least one of the following: clicks, orders, click-through followers, and online followers for the influencer's products corresponding to each reference keyword within a day. Click-through followers refer to the number of followers of the influencer's online followers who clicked on the influencer's products within a day.

[0070] Taking the example of a user who sells products and historical data for selling products, conversion data includes at least one of the following: click-through rate (CTR), conversion rate (CVR), and fan activity rate for each day in the historical time period. The click-through rate is positively correlated with the number of clicks and negatively correlated with the number of online fans. The conversion rate is positively correlated with the number of orders and negatively correlated with the number of clicks. The fan activity rate is the ratio of the number of fans who click to the number of fans who are online.

[0071] Taking the conversion data including the daily click-through rate, conversion rate and fan activity rate in the historical time period as an example, the average value of the click-through rate, conversion rate and fan activity rate can be determined as the conversion weight corresponding to the reference keyword.

[0072] In some embodiments, when there are multiple influencers, the click-through rate is positively correlated with the average click volume and negatively correlated with the average number of online followers. The average click volume is the average of the click volumes of multiple influencers. The average number of online followers is the average of the online followers of multiple influencers. The conversion rate is positively correlated with the average order volume and negatively correlated with the average click volume. The average order volume is the average of the order volumes of multiple influencers.

[0073] In the above embodiment, the conversion data is corrected using the average number of online fans, average number of clicks, and average number of orders corresponding to multiple users who bring goods, that is, the Bayesian smoothing idea is used for correction, which can make the conversion data closer to the true value, thereby further improving the accuracy of the recommendation.

[0074] return Figure 1 In step S40, information related to the keyword to be processed is recommended based on the similarity between the comprehensive vector and each target vector. For example, the similarity is the vector inner product or cosine similarity between the vectors.

[0075] If there are multiple objects, an object to be recommended is selected from the multiple objects, and relevant information about the object to be recommended is recommended. For example, a user to be recommended is selected from multiple users who are selling products, and the user information of the user to be recommended is recommended. In this case, a user matching the keywords to be processed is selected from the multiple users, so that the selected user can sell the items corresponding to the keywords to be processed. In some embodiments, the user information of the user corresponding to the comprehensive vector with the largest sum of similarities with each target vector can be recommended.

[0076] In the case where the object is one object, there are multiple target keywords. From the multiple target keywords, a keyword to be recommended is selected, and related information of the keyword to be recommended is recommended. For example, from multiple candidate items, an item to be recommended is selected, and the item information of the item to be recommended is recommended. In this case, a matching item to be recommended is selected for a certain user who is promoting the product, so that the user can promote the product for the item to be recommended. In some embodiments, the item information of the candidate item corresponding to the target vector with the greatest similarity to the comprehensive vector can be recommended.

[0077] In the above embodiment, the comprehensive vector of the object is determined by converting the keywords to be processed into target keywords and using the reference vectors of multiple reference keywords that belong to the same keyword category as the target keywords obtained based on historical data. Recommendations are made based on the similarity between the comprehensive vector and the target vector. The matching degree between the object and the keywords to be processed is comprehensively considered, which can improve the accuracy of the recommendation.

[0078] Figure 4 is a block diagram illustrating an information recommendation apparatus according to some embodiments of the present disclosure.

[0079] like Figure 4 As shown, the information recommendation device 4 includes a first determination module 41 , an acquisition module 42 , a second determination module 43 and a recommendation module 44 .

[0080] The first determining module 41 is configured to determine at least one target keyword and its target vector according to the keyword to be processed, for example, Figure 1 In step S10 shown in FIG, target keywords correspond to target vectors one by one.

[0081] The acquisition module 42 is configured to acquire, for each target keyword, a plurality of reference keywords belonging to the same keyword category as each target keyword from the historical data corresponding to the object related to the keyword to be processed, for example, by performing the following steps: Figure 1 Step S20 is shown.

[0082] The second determining module 43 is configured to determine a comprehensive vector corresponding to the object based on a plurality of reference vectors corresponding to a plurality of reference keywords, for example, by performing the following steps: Figure 1 In step S30 shown, the reference keywords correspond to the reference vectors one by one.

[0083] The recommendation module 44 is configured to recommend information related to the keyword to be processed based on the similarity between the comprehensive vector and each target vector, for example, Figure 1 Step S40 is shown.

[0084] Figure 5 is a block diagram illustrating an information recommendation device according to some other embodiments of the present disclosure.

[0085] like Figure 5 As shown, the information recommendation device 5 includes a memory 51 and a processor 52 coupled to the memory 51. The memory 51 is used to store instructions for executing the information recommendation method according to the embodiment. The processor 52 is configured to execute the information recommendation method according to any of the embodiments of the present disclosure based on the instructions stored in the memory 51.

[0086] Figure 6 is a block diagram illustrating a computer system for implementing some embodiments of the present disclosure.

[0087] like Figure 6 As shown, the computer system 60 may be embodied as a general-purpose computing device. The computer system 60 includes a memory 610, a processor 620, and a bus 600 that connects the various system components.

[0088] The memory 610 may include, for example, a system memory, a non-volatile storage medium, and the like. The system memory may store, for example, an operating system, an application program, a boot loader, and other programs. The system memory may include a volatile storage medium, such as a random access memory (RAM) and / or a cache memory. The non-volatile storage medium may store, for example, instructions for executing at least one corresponding embodiment of the information recommendation method. Non-volatile storage media include, but are not limited to, disk storage, optical storage, flash memory, and the like.

[0089] The processor 620 can be implemented as a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, or as discrete hardware components such as discrete gates or transistors. Accordingly, each module, such as the judgment module and the determination module, can be implemented by a central processing unit (CPU) executing instructions in a memory that execute corresponding steps, or by dedicated circuits that execute corresponding steps.

[0090] The bus 600 may use any of a variety of bus architectures, including, but not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MCA) bus, and a Peripheral Component Interconnect (PCI) bus.

[0091] The computer system 60 may also include an input / output interface 630, a network interface 640, a storage interface 650, and the like. These interfaces 630, 640, and 650, as well as the memory 610 and the processor 620, may be connected via a bus 600. The input / output interface 630 provides a connection interface for input / output devices such as a display, mouse, and keyboard. The network interface 640 provides a connection interface for various networked devices. The storage interface 650 provides a connection interface for external storage devices such as floppy disks, USB flash drives, and SD cards.

[0092] Here, various aspects of the present disclosure are described with reference to flowcharts and / or block diagrams of methods, devices, and computer program products according to embodiments of the present disclosure. It should be understood that each block of the flowcharts and / or block diagrams, and combinations of blocks, can be implemented by computer-readable program instructions.

[0093] These computer-readable program instructions may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable device to produce a machine, so that the processor executes the instructions to produce means for implementing the functions specified in one or more blocks in the flowcharts and / or block diagrams.

[0094] These computer-readable program instructions may also be stored in a computer-readable memory, which cause the computer to operate in a specific manner to produce an article of manufacture, including instructions for implementing the functions specified in one or more blocks in the flowcharts and / or block diagrams.

[0095] The present disclosure can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects.

[0096] The accuracy of recommendations can be improved through the information recommendation method and device, and the computer-storable medium in the above-mentioned embodiments.

[0097] The information recommendation method and apparatus, and computer-storable medium according to the present disclosure have been described in detail. To avoid obscuring the concepts of the present disclosure, some details known in the art have been omitted. Based on the above description, those skilled in the art will fully understand how to implement the technical solutions disclosed herein.

Claims

1. An information recommendation method, comprising: Determine at least one target keyword and its target vector based on the keyword to be processed, where the target keyword and the target vector have a one-to-one correspondence; For each target keyword, obtain multiple reference keywords belonging to the same keyword category as each target keyword from historical data corresponding to an object related to the keyword to be processed, wherein the object is a product-influenced user or advertising media; Determining a comprehensive vector corresponding to the object based on a plurality of reference vectors corresponding to the plurality of reference keywords, comprising: determining, for the object, conversion data corresponding to each reference keyword based on corresponding historical data, and determining the comprehensive vector based on the plurality of reference vectors and their corresponding conversion data, wherein the reference keywords correspond to the reference vectors in a one-to-one manner, and the conversion data includes at least one of a click-through rate, a conversion rate, and a fan activity rate; Recommending information related to the keyword to be processed according to the similarity between the comprehensive vector and each target vector, including: when the object is multiple objects, selecting an object to be recommended from the multiple objects and recommending relevant information of the object to be recommended.

2. The information recommendation method according to claim 1, wherein: Determining the integrated vector according to the multiple reference vectors and their corresponding conversion data includes: Determine the corresponding conversion weight based on the conversion data corresponding to each reference keyword; A weighted operation is performed on the multiple reference vectors using multiple conversion weights to obtain the comprehensive vector.

3. The information recommendation method according to claim 2, wherein: The object is a product promotion user, and the historical data is historical product promotion data. The historical product promotion data of the product promotion user includes product promotion data for each day in a historical period. The daily product promotion data includes at least one of the number of clicks, orders, click-through fans, and online fans of the product promotion items corresponding to each reference keyword in one day. The click-through fans are the number of fans who click on the product promotion items among the online fans of the product promotion user in one day. The conversion data includes at least one of the daily click-through rate, conversion rate and fan activity rate within the historical time period, the click-through rate is positively correlated with the number of clicks and negatively correlated with the number of online fans, the conversion rate is positively correlated with the order volume and negatively correlated with the number of clicks, and the fan activity rate is the ratio between the number of clicking fans and the number of online fans.

4. The information recommendation method according to claim 3, wherein: There are multiple users who bring goods, the click-through rate is positively correlated with the average number of clicks and negatively correlated with the average number of online fans, the average number of clicks is the average of the number of clicks of the multiple users who bring goods, and the average number of online fans is the average of the number of online fans of the multiple users who bring goods; The conversion rate is positively correlated with the average order volume and negatively correlated with the average click volume. The average order volume is the average of the order volumes of the multiple users who bring goods.

5. The information recommendation method according to claim 3, wherein: Determining the appropriate conversion weights includes: The average value of the click-through rate, the conversion rate and the fan activity rate is determined as the conversion weight.

6. The information recommendation method according to claim 1, wherein: The historical data corresponding to the object includes at least one set of order data derived from the object, each set of order data includes item identifiers of multiple items, and obtaining, from the historical data corresponding to the object related to the keyword to be processed, multiple reference keywords belonging to the same keyword category as each target keyword includes: determining, based on each set of order data derived from the object, a plurality of item groups, each item group including item identifiers of two different items; determining, based on each set of order data derived from the object, an association weight value between two items corresponding to the item identifiers in each item group, wherein a greater association weight value indicates a stronger correlation between the two items; Constructing a graph model with item identifiers in the plurality of item groups as vertices and associated weight values ​​as edges; By using the graph model, a plurality of reference keywords belonging to the same keyword category as each target keyword are obtained from historical data corresponding to the object.

7. The information recommendation method according to claim 6, wherein: The order data also includes other order parameters corresponding to the item identifier. Using the graph model, obtaining multiple reference keywords belonging to the same keyword category as each target keyword from historical data corresponding to the object includes: Using a random walk algorithm, randomly walk the graph model until a stop condition is met, and obtain at least one item identifier as a reference item identifier; For each target keyword, an item identifier or other order parameter that belongs to the same keyword category as the target keyword and corresponds to the reference item identifier is selected from each set of order data derived from the object as a reference keyword.

8. The information recommendation method according to claim 6, wherein: The order data also includes the order date. Determining the association weight value between two items corresponding to the item identifiers in each item group includes: An association weight value between two items corresponding to the item identifiers in each item group is determined based on a difference between order dates corresponding to the two item identifiers in each item group, where the association weight value is negatively correlated with the difference between the order dates.

9. The information recommendation method according to claim 6, wherein: The order data also includes the order date. Based on the order data corresponding to each product-carrying user, multiple product-carrying groups are determined to include: Determine, based on each set of order data, at least one item set, each item set including multiple item identifiers arranged in order of order, wherein the difference between the latest order date and the earliest order date corresponding to the multiple item identifiers in the same item set is less than a date threshold; For each item set, two adjacent item identifiers are grouped into one item group to obtain the multiple item groups.

10. The information recommendation method according to claim 1, wherein: In the case where the object is one object, the at least one target keyword includes multiple target keywords, and the recommended information related to the keyword to be processed includes: In the case that the object is one object, a keyword to be recommended is selected from the multiple target keywords, and related information of the keyword to be recommended is recommended.

11. The information recommendation method according to claim 1, wherein: Determining at least one target keyword based on the keyword to be processed includes: The keyword to be processed is converted into at least one target keyword according to a keyword category corresponding to the keyword to be processed.

12. The information recommendation method according to claim 1, wherein: In the case where the object is a user who brings goods, the keyword categories include user identification category, item identification category, brand identification category, store identification category and category.

13. An information recommendation device, comprising: A first determination module is configured to determine at least one target keyword and its target vector based on the keyword to be processed, where the target keyword and the target vector have a one-to-one correspondence; An acquisition module is configured to acquire, for each target keyword, a plurality of reference keywords belonging to the same keyword category as each target keyword from historical data corresponding to an object related to the keyword to be processed, wherein the object is a product-influenced user or advertising media; A second determination module is configured to determine a comprehensive vector corresponding to the object based on a plurality of reference vectors corresponding to the plurality of reference keywords, including: for the object, determining conversion data corresponding to each reference keyword based on corresponding historical data, and determining the comprehensive vector based on the plurality of reference vectors and their corresponding conversion data, wherein the reference keywords correspond to the reference vectors in a one-to-one manner, and the conversion data includes at least one of a click-through rate, a conversion rate, and a fan activity rate; The recommendation module is configured to recommend information related to the keyword to be processed based on the similarity between the comprehensive vector and each target vector, including: when the object is multiple objects, selecting the object to be recommended from the multiple objects and recommending relevant information of the object to be recommended.

14. An information recommendation device, comprising: Memory; as well as A processor coupled to the memory, wherein the processor is configured to execute the information recommendation method according to any one of claims 1 to 12 based on instructions stored in the memory.

15. A computer storable medium having computer program instructions stored thereon, wherein when the instructions are executed by a processor, the information recommendation method according to any one of claims 1 to 12 is implemented.

Images