Method and device for pushing personalized content

Abstract

The embodiment of the present invention provides a personalized content push method and device, including: obtaining a user's shopping behavior data and transaction behavior data; determining the user's expected shopping data based on the shopping behavior data and transaction behavior data; processing the expected shopping data according to a preset personalized algorithm to obtain the user's preferred shopping data; and making personalized product recommendations to the user based on the preferred shopping data. The user's shopping behavior data and transaction behavior data are collected and processed to obtain the user's expected shopping data, and then the expected shopping data is parsed by a preset personalized algorithm to obtain the user's preferred shopping data. Based on the preferred shopping data, personalized product recommendations are made to the user, thereby achieving accurate recommendation of personalized products to the user, and improving the accuracy and efficiency of product recommendations.

Description

Technical Field

[0001] The present invention relates to the field of wireless charging, and in particular to a method and device for pushing personalized content. Background Art

[0002] In the field of e-commerce, personalized product recommendations have become one of the key technologies to improve user experience and increase sales conversion. With the rapid development of big data and artificial intelligence technologies, more and more e-commerce companies are trying to achieve more accurate personalized recommendations by analyzing users' shopping behavior data and transaction behavior data.

[0003] In the existing personalized content push method technology, since the traditional personalized content push method mainly relies on the user's historical purchase records and simple similarity matching algorithms, it is difficult to capture the user's complex and changeable shopping needs and preferences. Therefore, in the existing personalized content push method, there is a problem that it is impossible to make personalized product recommendations to users based on the user's historical purchase records and matching algorithms. Summary of the invention

[0004] An embodiment of the present invention provides a method for pushing personalized content, which collects and processes the user's shopping behavior data and transaction behavior data to obtain the user's expected shopping data, and then parses the expected shopping data through a preset personalized algorithm to obtain the user's preferred shopping data. Based on the analysis of the preferred shopping data, personalized product recommendations can be made to the user.

[0005] In a first aspect, an embodiment of the present invention provides a method for pushing personalized content, the method comprising the following steps:

[0006] Obtain users' shopping behavior data and transaction behavior data;

[0007] Determining the user's expected shopping data based on the shopping behavior data and the transaction behavior data;

[0008] Processing the desired shopping data according to a preset personalized algorithm to obtain the user's preferred shopping data;

[0009] Based on the preferred shopping data, personalized product recommendations are made to the user.

[0010] Optionally, the shopping behavior data includes historical behavior data and current shopping behavior data, the transaction behavior data includes historical transaction behavior data, and the acquiring of the user's shopping behavior data and transaction behavior data includes:

[0011] When a user is browsing a current product, querying the user's historical shopping behavior data for the current product, and detecting the user's current shopping behavior data when browsing the current product;

[0012] The historical transaction behavior data of the user for the current commodity is queried to obtain the transaction behavior data of the user.

[0013] Optionally, determining the user's expected shopping data based on the shopping behavior data and the transaction behavior data includes:

[0014] Determining first desired shopping data of the user based on the historical shopping behavior data and the current shopping behavior data;

[0015] Determining second desired shopping data of the user based on the historical transaction behavior data;

[0016] Based on the first expected shopping data and the second expected shopping data, the user's expected shopping data is determined, and the expected shopping data is used to judge the possibility of the user purchasing the product.

[0017] Optionally, the processing the expected shopping data according to a preset personalized algorithm to obtain the user's preferred shopping data includes:

[0018] Performing expected purchase prediction processing on the expected shopping data by using the preset personalized algorithm to obtain personalized interest data of the user, wherein the personalized interest data includes the user's interest value for the current product;

[0019] Based on the user's interest value for the current commodity, the user's preferred shopping data is determined.

[0020] Optionally, the determining the first expected shopping data of the user based on the historical shopping behavior data and the current shopping behavior data, the method further comprises:

[0021] Performing similarity processing on the current shopping behavior and the user's historical shopping behavior;

[0022] If the similarity between the current shopping behavior and the historical shopping behavior is less than a preset similarity threshold, the current shopping behavior is used as the historical shopping behavior of the user;

[0023] If the similarity between the current shopping behavior and the historical shopping behavior is greater than or equal to a preset similarity threshold, the historical shopping behavior of the user is retained.

[0024] Optionally, the method further comprises: performing personalized product recommendations for the user based on the preferred shopping data:

[0025] Based on the current shopping behavior habits and current shopping interest data, matching associated users who have the same shopping behavior habits and shopping interest data as the user;

[0026] Acquire the associated user's preferred shopping data, wherein the preferred shopping data includes content features of preferred products and shopping behavior habits for the preferred products;

[0027] Based on the content features of the associated user's preferred products and the shopping behavior habits for the preferred products, a target product that is within a product similarity threshold interval with the preferred products is calculated, and the target product is pushed to the user.

[0028] In a second aspect, an embodiment of the present invention further provides a personalized content push device, wherein the personalized content push method device comprises:

[0029] An acquisition module is used to acquire the user's shopping behavior data and transaction behavior data;

[0030] A determination module, used to determine the user's expected shopping data based on the shopping behavior data and the transaction behavior data;

[0031] A processing module, used to process the expected shopping data according to a preset personalized algorithm to obtain the user's preferred shopping data;

[0032] A recommendation module is used to make personalized product recommendations to the user based on the preferred shopping data.

[0033] In a third aspect, an embodiment of the present invention further provides a personalized content push system, wherein the personalized content push method system comprises: a personalized content push method device, a server, and a client.

[0034] In a fourth aspect, an embodiment of the present invention provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps in the personalized content push method provided in the embodiment of the present invention when executing the computer program.

[0035] In a fifth aspect, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps in the method for pushing personalized content provided in the embodiment of the invention are implemented.

[0036] In the embodiment of the present invention, the user's shopping behavior data and transaction behavior data are obtained; based on the shopping behavior data and transaction behavior data, the user's expected shopping data is determined; the expected shopping data is processed according to a preset personalized algorithm to obtain the user's preferred shopping data; based on the preferred shopping data, personalized product recommendations are made to the user. The user's expected shopping data is obtained by collecting the user's shopping behavior data and transaction behavior data, and then the expected shopping data is parsed by a preset personalized algorithm to obtain the user's preferred shopping data. Based on the preferred shopping data, personalized product recommendations are made to the user, thereby achieving accurate recommendation of personalized products to the user, and improving the accuracy and efficiency of product recommendations. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0038] Figure 1 This is an architecture diagram of a personalized content push system provided by an embodiment of the present invention;

[0039] Figure 2 is a flow chart of a method for pushing personalized content provided by an embodiment of the present invention;

[0040] Figure 3 yes Figure 2 Detailed flow chart of step 201;

[0041] Figure 4 yes Figure 2 Detailed flow chart of step 202;

[0042] Figure 5 yes Figure 2 Detailed flow chart of step 203;

[0043] Figure 6 yes Figure 4 Detailed flowchart of step 2021;

[0044] Figure 7 yes Figure 2 Detailed flow chart of step 204;

[0045] Figure 8 It is a structural schematic diagram of a personalized content push device provided in an embodiment of the present invention;

[0046] Fig. 9It is a structural schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] like Figure 1 As shown, Figure 1 1 is an architecture diagram of a personalized content push system 100 provided in an embodiment of the present invention, and the personalized content push system 100 includes: a personalized content push device 101, a server 102, and a client 103. The personalized content push device 800 further includes: an acquisition module for acquiring the user's shopping behavior data and transaction behavior data; a determination module for determining the user's expected shopping data based on the shopping behavior data and transaction behavior data; a processing module for processing the expected shopping data according to a preset personalized algorithm to obtain the user's preferred shopping data; and a recommendation module for making personalized product recommendations to the user based on the preferred shopping data.

[0049] The shopping behavior data may include but are not limited to historical shopping behavior data and current shopping behavior data. Specifically, the historical shopping behavior data may include but are not limited to the user's purchase history data, browsing history data, search history data, collection and preference data on the shopping page, and user feedback data on the product. The shopping page may be a built-in shopping page corresponding to an APP equipped with the personalized content push system, or any other page and electronic device that can be used for shopping.

[0050] The above-mentioned transaction behavior data may include but are not limited to historical transaction behavior data. Specifically, the above-mentioned historical transaction behavior data may include but are not limited to the user's payment record data, the corresponding order data of the purchased goods, and the payment behavior data, etc. It should be noted that the above-mentioned payment behavior data may include but are not limited to data such as successful payment, failed payment, payment method and payment frequency. The above-mentioned payment behavior data can also be used to determine the user's operating behavior during the transaction process.

[0051] More specifically, the shopping behavior data and transaction behavior data can be obtained by tracking the user's interaction on the purchase page through the personalized content push system. Generally speaking, the user's behavior on the website or APP (such as clicks, searches, purchases, etc.) can be tracked to collect the current user's shopping behavior data, and the current user's previous historical shopping data can be obtained by tracking logs. Through the log data of the corresponding products in the historical shopping data in the database, the user's corresponding transaction behavior data can also be tracked.

[0052] The above-mentioned expected shopping data can be used to determine the possibility of the user purchasing the currently browsed product, that is, the more ideal the expected shopping data is, the higher the user's desire to purchase the currently browsed product is, and the more likely the user is to purchase the currently browsed product. Specifically, the above-mentioned expected shopping data can be obtained by analyzing the products with the same attributes as the currently browsed product in the user's historical browsed products through the above-mentioned personalized content push system, and can also be obtained by analyzing the user's facial data when browsing the current product, so as to determine the user's emotions when browsing the current product, and thus determine whether the user has a purchase tendency for the currently browsed product.

[0053] More specifically, the above-mentioned expected shopping data can also be used to analyze the user's shopping behavior data and transaction behavior data to determine whether there is a purchasing tendency for goods with the same attributes, that is, whether the expected shopping data corresponding to the goods with the same attributes is ideal, so that the corresponding goods can be pushed according to the user's expected shopping data.

[0054] The above-mentioned preset personalized algorithm can be applied to a deep learning model that can parse the above-mentioned expected shopping data to obtain the user's corresponding preference data. Specifically, the above-mentioned preset personalized algorithm may include but is not limited to collaborative filtering algorithms, content-based recommendation algorithms, hybrid recommendation algorithms, and context-aware recommendation algorithms. It should be noted that the above-mentioned algorithm can be adaptively adjusted according to the above-mentioned personalized content push system. For example, the user's shopping behavior data and transaction behavior data can be added to the above-mentioned preset personalized algorithm to adjust the algorithm's input parameters. It is understandable that the above-mentioned deep learning model may include but is not limited to convolutional neural network models, multi-layer perceptron models, and recurrent neural networks (RNNs) and other deep learning models that can be equipped with the above-mentioned preset personalized algorithm.

[0055] The above-mentioned preferred shopping data may include, but are not limited to, content features of preferred products and shopping behavior habits for preferred products. Specifically, the above-mentioned preferred shopping data may reflect the user's preference for a certain type or corresponding product with a certain product attribute. In a possible embodiment, the above-mentioned personalized content push system may parse the user's preferred shopping data to obtain the user's preference for a certain category of products in the current time period.

[0056] Specifically, the content features of the preferred products may include but are not limited to the corresponding commodity attributes such as the price range, usage scenarios, categories, and commodity characteristics of the preferred products. Generally speaking, users have a certain desire to buy a certain commodity because they are attracted by certain characteristics of the commodity. Therefore, the personalized content push system can analyze the content features of the preferred products to derive the characteristics of the preferred products, and finally push other commodities with the same commodity attributes to the user based on the characteristics of the preferred products.

[0057] The above-mentioned shopping behavior habits for preferred products can be used to determine the user's processing habits for preferred products. For example, for preferred products, the user may have a habit of collecting them or directly purchasing them. Based on the shopping behavior habits, the user may be prompted for products with the same product attributes according to the same shopping behavior habits, so as to facilitate the user to directly operate.

[0058] In a possible embodiment, the personalized product recommendations can be made by analyzing the user's shopping behavior habits for preferred products through the personalized content push system, and the corresponding products with the same product attributes can be pushed based on the analysis results. The products with the same product attributes can also be processed with corresponding shopping behavior habits based on the analysis results.

[0059] In this embodiment, the personalized content push system collects and processes the user's shopping behavior data and transaction behavior data to obtain the user's expected shopping data, and then parses the expected shopping data through a preset personalized algorithm to obtain the user's preferred shopping data. Based on the analysis of the preferred shopping data, personalized product recommendations can be made to the user.

[0060] In this personalized content push system, the user participation can also be improved through interactive activities, member benefits and other content according to the user participation improvement module, to stimulate the user's enthusiasm for participation and increase user activity. Specifically, the personalized content push system can collect and analyze user feedback, determine user needs and problems, and adjust the system's functions and push strategies, so that the personalized content push system can push products to users more accurately.

[0061] The above-mentioned user engagement enhancement module can also improve user stickiness by designing user engagement enhancement strategies and implementing engagement enhancement strategies. Specifically, the above-mentioned engagement enhancement strategies may include but are not limited to designing interactive activities, designing member-exclusive benefits, and designing community building activities. More specifically, various interactive activities such as sweepstakes, online discussions, etc. can be designed to stimulate user enthusiasm for participation. Member-exclusive discount activities, gift redemption and other benefits can also be provided to increase user motivation for participation. An active member community platform can also be established to encourage interaction and communication between users and improve user engagement. Finally, the system will analyze the corresponding data on user engagement to obtain strategies to enhance user engagement, thereby ensuring user activity and loyalty.

[0062] like Figure 2 As shown, Figure 2 2 is a flow chart of a personalized content push method provided by an embodiment of the present invention. The personalized content push method includes steps 201-204, wherein:

[0063] 201. Obtain the user's shopping behavior data and transaction behavior data.

[0064] In an embodiment of the present invention, the above-mentioned personalized content push method can be applied to the above-mentioned personalized content push method system. The above-mentioned personalized content push system has service functions such as personalized product push processing and personalized product customization, and can be constructed based on a server or a server cluster. The above-mentioned server or server cluster can be an electronic device with personalized product push capability.

[0065] The shopping behavior data may include but are not limited to historical shopping behavior data and current shopping behavior data. Specifically, the historical shopping behavior data may include but are not limited to the user's purchase history data, browsing history data, search history data, collection and preference data on the shopping page, and user feedback data on the product. The shopping page may be a built-in shopping page corresponding to an APP equipped with the personalized content push system, or any other page and electronic device that can be used for shopping.

[0066] The above-mentioned transaction behavior data may include but are not limited to historical transaction behavior data. Specifically, the above-mentioned historical transaction behavior data may include but are not limited to the user's payment record data, the corresponding order data of the purchased goods, and the payment behavior data, etc. It should be noted that the above-mentioned payment behavior data may include but are not limited to data such as successful payment, failed payment, payment method and payment frequency. The above-mentioned payment behavior data can also be used to determine the user's operating behavior during the transaction process.

[0067] In a possible embodiment, the shopping behavior data and the transaction behavior data can be obtained by tracking the user's interaction on the purchase page through the personalized content push system. Generally speaking, the user's behavior on the website or APP (such as clicks, searches, purchases, etc.) can be tracked to collect the current user's shopping behavior data, and the current user's previous historical shopping data can be obtained by tracking logs. Through the log data of the corresponding goods in the historical shopping data in the database, the user's corresponding transaction behavior data can also be tracked.

[0068] 202. Determine the user's expected shopping data based on the shopping behavior data and the transaction behavior data.

[0069] In the embodiment of the present invention, the above-mentioned expected shopping data can be used to judge the possibility of the user purchasing the currently browsed product, that is, the more ideal the expected shopping data is, the higher the user's desire to purchase the currently browsed product is, and the more likely the user is to purchase the currently browsed product. Specifically, the above-mentioned expected shopping data can be obtained by analyzing the products with the same attributes as the currently browsed product in the user's historical browsed products through the above-mentioned personalized content push system, and can also be obtained by analyzing the user's facial data when browsing the current product, so as to judge the user's emotions when browsing the current product, and thus judge whether the user has a purchase tendency for the currently browsed product.

[0070] In a possible embodiment, the personalized content push system can analyze the user's shopping behavior data and transaction behavior data to determine whether the user has a purchasing tendency for goods with the same attributes, that is, whether the expected shopping data corresponding to the goods with the same attributes is ideal, so that the corresponding goods can be pushed according to the user's expected shopping data.

[0071] 203. Process the expected shopping data according to a preset personalized algorithm to obtain the user's preferred shopping data.

[0072] In an embodiment of the present invention, the above-mentioned preferred shopping data may include but is not limited to content features of preferred products and shopping behavior habits for preferred products. Specifically, the above-mentioned preferred shopping data may reflect the user's preference for a certain type or corresponding product with a certain commodity attribute. In a possible embodiment, the above-mentioned personalized content push system can parse the user's preferred shopping data to obtain the user's preference for a certain category of goods in the current time period.

[0073] In a possible embodiment, the personalized content push system may analyze the product attributes corresponding to the price range, usage scenarios, categories, and product features of the preferred products, thereby deriving the characteristics of the preferred products, and finally push other products with the same product attributes to the user based on the characteristics of the preferred products.

[0074] 204. Provide personalized product recommendations to users based on their shopping preference data.

[0075] In an embodiment of the present invention, the above-mentioned shopping behavior habits for preferred products can be used to determine the user's processing habits for preferred products. For example, for preferred products, the user may have a habit of collecting them or directly purchasing them. Based on the shopping behavior habits, the user may be prompted for products with the same product attributes according to the same shopping behavior habits, so as to facilitate the user to directly operate.

[0076] In a possible embodiment, the personalized product recommendations can be made by analyzing the user's shopping behavior habits for preferred products through the personalized content push system, and the corresponding products with the same product attributes can be pushed based on the analysis results. The products with the same product attributes can also be processed with corresponding shopping behavior habits based on the analysis results.

[0077] In the embodiment of the present invention, the user's shopping behavior data and transaction behavior data are obtained; based on the shopping behavior data and transaction behavior data, the user's expected shopping data is determined; the expected shopping data is processed according to a preset personalized algorithm to obtain the user's preferred shopping data; based on the preferred shopping data, personalized product recommendations are made to the user. The user's expected shopping data is obtained by collecting the user's shopping behavior data and transaction behavior data, and then the expected shopping data is parsed by a preset personalized algorithm to obtain the user's preferred shopping data. Based on the preferred shopping data, personalized product recommendations are made to the user, thereby achieving accurate recommendation of personalized products to the user, and improving the accuracy and efficiency of product recommendations.

[0078] Optional, such as Figure 3 As shown, in step 201, the step of obtaining the user's shopping behavior data and transaction behavior data also includes 2011-2012, wherein:

[0079] 2011. When the user is browsing the current product, query the user's historical shopping behavior data for the current product, and detect the user's current shopping behavior data when browsing the current product.

[0080] In an embodiment of the present invention, the above-mentioned current product can be determined by the user's operation on the shopping page. Generally speaking, when a user clicks on a shopping page to enter the browsing page of a certain product, the above-mentioned personalized content push system will retrieve the information of the current product that the user is browsing through the background database to obtain the information of the currently browsed product.

[0081] The historical shopping behavior data of the user when browsing the current product and the corresponding current shopping behavior data can also be detected in the log data by using the retrieved product information, including but not limited to the browsing time of the current product, whether to click on the collection, etc. It should be noted that if the database does not contain the historical shopping behavior data of the user for the currently browsed product, the current shopping behavior data will be entered into the historical shopping behavior data for the current product, and when the current shopping behavior data changes, the changed current shopping behavior data will be entered into the historical shopping behavior data of the currently browsed product in real time.

[0082] In a possible embodiment, the personalized content push system may acquire the historical shopping behavior data of the user when browsing the current product in real time, and acquire the corresponding current shopping behavior data in real time.

[0083] 2012. Query the historical transaction behavior data of the user for the current product to obtain the transaction behavior data of the user.

[0084] In an embodiment of the present invention, the personalized content push system can also query the historical transaction behavior data of the current product to determine whether the user has purchased the product, and can judge whether the user has a purchase tendency for the product based on whether the user has purchased the product. Specifically, in the process of querying the historical transaction behavior data, it is also necessary to query whether there is a return or refund record. When there is a return or refund record, a new judgment will be made when judging the purchase tendency.

[0085] Optional, such as Figure 4 As shown, in step 202, based on the shopping behavior data and the transaction behavior data, the user's expected shopping data is determined, and also includes 2021-2023, wherein:

[0086] 2021. Determine the user's first expected shopping data based on historical shopping behavior data and current shopping behavior data.

[0087] In an embodiment of the present invention, the first expected shopping data may be a result obtained by comprehensively analyzing the historical shopping behavior data and the current shopping behavior data through the personalized content push system. The first expected shopping data may be used to analyze whether the user has a purchase tendency for the current product or to judge the degree of preference for the current product.

[0088] In a possible embodiment, fitting data is obtained by fitting the data of goods that the user has purchased in the past, including the time of purchase, purchase quantity, purchase frequency, the number of times the corresponding goods have been viewed, the browsing duration, and the user's evaluation of the purchased goods, and the current real-time shopping behavior data is detected, and the current shopping behavior data is comprehensively analyzed with the above-mentioned fitting data to determine the user's first expected shopping data for the current product.

[0089] 2022. Determine the user's second expected shopping data based on historical transaction behavior data.

[0090] The second expected shopping data may be a result obtained by analyzing the historical transaction behavior data through the personalized content push system. The second expected shopping data may be used in combination with the first expected shopping data to analyze whether the user has a purchase tendency for the current product or to judge the degree of preference for the current product.

[0091] According to the above steps, whether the user has a purchase intention for the product can be further determined through the historical transaction behavior data of the current product.

[0092] 2023. Determine the user's expected shopping data based on the first expected shopping data and the second expected shopping data.

[0093] The above-mentioned expected shopping data can be used to determine the possibility of the user purchasing the product. Generally speaking, the closer the expected shopping data is to the shopping threshold, the greater the user's desire to purchase the current product can be determined. The above-mentioned shopping threshold can be set by the first expected shopping data and the second expected shopping data. Generally speaking, the more change parameters there are in the first expected shopping data and the second expected shopping data, the larger the shopping threshold can be set. It should be noted that the specific value of the above-mentioned shopping threshold can be determined according to the specific implementation plan. The above-mentioned change parameters can be obtained by detecting the user's shopping behavior operations. Generally speaking, the longer the user stays on the page of a certain product, or the more times the user clicks, the more change parameters there are.

[0094] In a possible embodiment, the personalized content push system obtains the user's first expected shopping data for the current product by analyzing the user's historical shopping behavior and current shopping behavior, and obtains the user's historical transaction behavior data for the current product by analyzing the user's historical transaction behavior data. It should be noted that if the user has no purchase record or browsing record for the current product, the shopping threshold is set to the lowest, and the user's shopping tendency for the current product is judged based on the subsequent user's behavior data on the shopping page of the current product.

[0095] Optional, such as Figure 5 As shown, in step 203, the expected shopping data is processed according to a preset personalized algorithm to obtain the user's preferred shopping data, and also includes 2031-2032, wherein:

[0096] 2031. Perform expected purchase prediction processing on the expected shopping data through a preset personalized algorithm to obtain the user's personalized interest data.

[0097] The preset personalized algorithm can be applied to the expected purchase prediction processing and analysis of the expected shopping data, so as to obtain a deep learning model of personalized interest data corresponding to the user. Specifically, the personalized interest data may include but is not limited to the user's interest value for the current product. Generally speaking, the larger the interest value, the higher the user's desire to buy the current product, and the more likely the user is to buy the current product.

[0098] The above-mentioned preset personalized algorithms may include but are not limited to collaborative filtering algorithms, content-based recommendation algorithms, hybrid recommendation algorithms, and context-aware recommendation algorithms. It should be noted that the above-mentioned algorithms can be adaptively adjusted according to the above-mentioned personalized content push system. For example, the user's shopping behavior data and transaction behavior data can be added to the above-mentioned preset personalized algorithms to adjust the input parameters of the algorithm. It is understandable that the above-mentioned deep learning models may include but are not limited to convolutional neural network models, multi-layer perceptron models, and recurrent neural networks (RNNs) and other deep learning models that can be equipped with the above-mentioned preset personalized algorithms.

[0099] In this embodiment, a collaborative filtering algorithm based on user behavior data and similarities between users can be used to recommend products or services that are similar to the user's preferences and that are liked by other users. A content recommendation algorithm based on the user's historical behavior and the characteristics of the product content can also be used to recommend products or services related to the user's interests.

[0100] Specifically, the collected user desired shopping data is cleaned and preprocessed through the personalized content push system to construct a user-product matrix, and other users with similar favorite products are searched in the system's affiliated database based on the user-product matrix, and the products liked by other users are recommended to the current user.

[0101] 2032. Determine the user's preferred shopping data based on the user's interest value for the current product.

[0102] In this embodiment, the user's preferred shopping data of different degrees can be determined by dividing the interest value into intervals. Generally speaking, the interest value can be divided into three intervals of "not interested", "somewhat interested" and "very interested", and the interval of the user's interest value for the current product can be determined by assigning corresponding values. The calculation of the above interest value can be obtained by calculating the expected shopping data through the above preset personalized algorithm.

[0103] The above-mentioned preferred shopping data may include but are not limited to content features of preferred products and shopping behavior habits for preferred products. Specifically, the above-mentioned preferred shopping data may reflect the user's preference for corresponding products of a certain type or with certain product attributes.

[0104] In a possible embodiment, the personalized content push system may analyze the user's preferred shopping data to obtain the user's preference for a certain category of goods in the current time period.

[0105] Optional, such as Figure 6 As shown, in step 2021, based on the historical shopping behavior data and the current shopping behavior data, the user's first expected shopping data is determined, and also includes 20211-20213, wherein:

[0106] 20211. Perform similarity processing on the current shopping behavior and the user’s historical shopping behavior.

[0107] In an embodiment of the present invention, the similarity between the user's historical shopping behavior and current shopping behavior can also be calculated through the above-mentioned personalized content push system to obtain a similarity value between the two, and whether there is a correlation between the user's historical shopping behavior and the current shopping behavior can be determined based on the similarity value. If there is a correlation, it means that the shopping tendency for the current product corresponding to the current shopping behavior is consistent with the product corresponding to the historical shopping behavior to a certain extent. Therefore, it can be judged whether the user has a purchasing tendency for the current product through the behavior analysis results of the products corresponding to the historical shopping behavior among the products purchased by the user.

[0108] 20212. If the similarity between the current shopping behavior and the historical shopping behavior is less than a preset similarity threshold, the current shopping behavior is taken as the user's historical shopping behavior.

[0109] The preset similarity threshold can be used to determine the correlation between the current shopping behavior and the historical shopping behavior. Specifically, the similarity threshold can also be set according to the number of change parameters between the current shopping behavior and the historical shopping behavior. The specific threshold interval can be determined according to the specific implementation plan. Generally speaking, the change parameters can be obtained by detecting the user's shopping behavior operations. Generally speaking, the longer the user stays on the page of a certain product, or the more times the user clicks, the more change parameters there are.

[0110] In this embodiment, when the similarity between the current shopping behavior and the historical shopping behavior is less than the above-mentioned preset similarity threshold, it can be explained that there is a certain degree of difference between the current shopping behavior and the user's historical shopping behavior. This difference can be used to judge that there is a huge difference between the user's shopping tendency for the current product and the product corresponding to the historical shopping. It can be understood that in this case, when the current shopping behavior is judged to have no shopping tendency, it can be explained that the historical shopping behavior can be said to have a shopping tendency. At this time, it means that the product corresponding to the historical shopping behavior is liked by the user, while the product corresponding to the current shopping behavior is not liked by the user.

[0111] It should be noted that the above-mentioned personalized content push system can treat the current shopping behavior as a new historical shopping behavior and store it in the user's corresponding database when the similarity between the current shopping behavior and the historical shopping behavior is less than a preset similarity threshold, so as to ensure that accurate personalized product push can be provided to the user in the future.

[0112] 20213. If the similarity between the current shopping behavior and the historical shopping behavior is greater than or equal to a preset similarity threshold, the user's historical shopping behavior is retained.

[0113] In this embodiment, when the similarity between the current shopping behavior and the historical shopping behavior is greater than or equal to the above-mentioned preset similarity threshold, it can be explained that the current shopping behavior is consistent with the user's historical shopping behavior, that is, it is explained that the user's purchase tendency for the current product is consistent with the purchase tendency for the product corresponding to the historical shopping behavior. It can be understood that in this case, when the current shopping behavior is judged to have no shopping tendency, it can be explained that the historical shopping behavior may also be without shopping tendency. At this time, it means that the product corresponding to the historical shopping behavior is not liked by the user, and the product corresponding to the current shopping behavior is also not liked by the user.

[0114] It should be noted that the above-mentioned personalized content push system can retain the original historical shopping behavior in the user's corresponding database when the similarity between the current shopping behavior and the historical shopping behavior is greater than or equal to the preset similarity threshold, so as to ensure that accurate personalized product push can be performed for the user in the future.

[0115] Optional, such as Figure 7 As shown, in step 204, personalized product recommendations are made to the user based on the preferred shopping data, and also include 2041-2043, wherein:

[0116] 2041. Based on the current shopping behavior habits and current shopping interest data, match associated users who have the same shopping behavior habits and shopping interest data as the user.

[0117] In an embodiment of the present invention, the current shopping behavior habits can be obtained by detecting the user's browsing behavior, clicking behavior, searching behavior, purchasing behavior and other shopping behavior data for the current product through the personalized content push system. Specifically, the data can be obtained by the system monitoring the page operations during the user's browsing of the product page. More specifically, by systematically collecting and analyzing these current shopping behavior habits, the effect of the recommendation system can be greatly improved, making the recommendations more accurate and personalized. The current shopping behavior habits can also be collected and obtained through the front-end behavior tracking technology of the website or application (such as cookies and JavaScript tracking scripts).

[0118] The above current shopping interest data can reflect the user's attention and preference for goods or services at a specific time point. Such data is critical for understanding the user's immediate needs and providing personalized recommendations. Specifically, the above current shopping interest data can be used to calculate the user's interest value for the current product. Generally speaking, the higher the interest value, the higher the user's purchase tendency for the current product.

[0119] In a possible embodiment, the personalized content push system matches the associated users who have the same shopping behavior habits and shopping interest data as the user in a database attached to the system through the user's current shopping behavior habits and current shopping interest data, and queries the associated users' preferred item information in the corresponding database.

[0120] 2042. Obtain the preferred shopping data of the associated user.

[0121] The preferred item information of the above-mentioned associated users is extracted through the above-mentioned personalized content push system to obtain the shopping data of the above-mentioned associated users, including but not limited to their preferred product content features (such as brand, price range, commodity classification, etc.) and shopping behavior habits (such as purchase time, purchase frequency, etc.)

[0122] 2043. Based on the content features of the associated user's preferred products and the shopping behavior habits for the preferred products, calculate the target products that are within the product similarity threshold interval with the preferred products, and push the target products to the user.

[0123] In an embodiment of the present invention, the personalized content push system can find other products similar to the associated user's preferred products through a content similarity algorithm (such as cosine similarity). These products need to meet a certain similarity threshold to ensure that they match the user's preferences. These target products that meet the similarity threshold are recommended to the current user. Among them, the recommended products can be displayed to the user through email marketing, in-app notifications, or personalized homepage display.

[0124] Specifically, the above-mentioned product similarity threshold interval can be set according to the number of associated users and the number of preferred items. The specific setting interval can be determined by the specific implementation plan. Generally speaking, the more the number of associated users and the corresponding number of preferred items, the larger the set product similarity threshold interval can be, so as to ensure that the corresponding personalized product content can be accurately pushed to the user.

[0125] like Figure 8 As shown, the embodiment of the present invention further provides a personalized content push method device 800, and the personalized content push method device includes:

[0126] Acquisition module 801, used to acquire the user's shopping behavior data and transaction behavior data;

[0127] A determination module 802, configured to determine the user's expected shopping data based on the shopping behavior data and the transaction behavior data;

[0128] A processing module 803 is used to process the expected shopping data according to a preset personalized algorithm to obtain the user's preferred shopping data;

[0129] The recommendation module 804 is used to make personalized product recommendations to the user based on the preferred shopping data.

[0130] Optionally, the acquisition module 801 includes:

[0131] A query submodule, used to query the user's historical shopping behavior data for the current product when the user is browsing the current product, and detect the user's current shopping behavior data when browsing the current product;

[0132] The first acquisition submodule is used to query the historical transaction behavior data of the user for the current commodity to obtain the transaction behavior data of the user.

[0133] Optionally, the determination module 802 includes:

[0134] A first determination submodule, configured to determine the user's first expected shopping data based on the historical shopping behavior data and the current shopping behavior data;

[0135] A second determination submodule, configured to determine the user's second expected shopping data based on the historical transaction behavior data;

[0136] The third determination submodule is used to determine the user's expected shopping data based on the first expected shopping data and the second expected shopping data, and the expected shopping data is used to judge the possibility of the user purchasing the product.

[0137] Optionally, the processing module 803 includes:

[0138] A second acquisition submodule is used to perform expected purchase prediction processing on the expected shopping data by using the preset personalized algorithm to obtain personalized interest data of the user, wherein the personalized interest data includes the user's interest value for the current product;

[0139] The fourth determination submodule is used to determine the user's preferred shopping data based on the user's interest value for the current product.

[0140] Optionally, the first determining submodule includes:

[0141] A first processing submodule, configured to perform similarity processing between the current shopping behavior and the user's historical shopping behavior;

[0142] A first shopping submodule, configured to use the current shopping behavior as the user's historical shopping behavior if the similarity between the current shopping behavior and the historical shopping behavior is less than a preset similarity threshold;

[0143] The second shopping submodule is used to retain the user's historical shopping behavior if the similarity between the current shopping behavior and the historical shopping behavior is greater than or equal to a preset similarity threshold.

[0144] Optionally, the recommendation module 804 includes:

[0145] A first matching submodule, for matching associated users who have the same shopping behavior habits and shopping interest data as the user based on the current shopping behavior habits and current shopping interest data;

[0146] A third acquisition submodule is used to acquire the associated user's preferred shopping data, wherein the preferred shopping data includes content features of preferred products and shopping behavior habits for preferred products;

[0147] The recommendation submodule is used to calculate the target product that is within the product similarity threshold interval with the preferred product based on the content features of the associated user's preferred product and the shopping behavior habits for the preferred product, and push the target product to the user.

[0148] like Fig. 9 As shown, an embodiment of the present invention further provides an electronic device 900, including a processor, and the processor can execute any one of the above-mentioned personalized content push methods.

[0149] Specifically, it includes a processor 901 and a memory 902, and a computer program for executing a method for pushing personalized content that is stored in the memory 902 and can be run on the processor 901, wherein:

[0150] The processor 901 runs the computer program of the personalized content push method stored in the memory 902, and performs the following steps:

[0151] Obtain users' shopping behavior data and transaction behavior data;

[0152] Determining the user's expected shopping data based on the shopping behavior data and the transaction behavior data;

[0153] Processing the desired shopping data according to a preset personalized algorithm to obtain the user's preferred shopping data;

[0154] Based on the preferred shopping data, personalized product recommendations are made to the user.

[0155] Optionally, the processor 901 executes the shopping behavior data including historical shopping behavior data and current shopping behavior data, the transaction behavior data including historical transaction behavior data, and the acquiring of the user's shopping behavior data and transaction behavior data includes:

[0156] When a user is browsing a current product, querying the user's historical shopping behavior data for the current product, and detecting the user's current shopping behavior data when browsing the current product;

[0157] The historical transaction behavior data of the user for the current commodity is queried to obtain the transaction behavior data of the user.

[0158] Optionally, the processor 901 performs the step of determining the user's expected shopping data based on the shopping behavior data and the transaction behavior data, including:

[0159] Determining first desired shopping data of the user based on the historical shopping behavior data and the current shopping behavior data;

[0160] Determining second desired shopping data of the user based on the historical transaction behavior data;

[0161] Based on the first expected shopping data and the second expected shopping data, the user's expected shopping data is determined, and the expected shopping data is used to judge the possibility of the user purchasing the product.

[0162] Optionally, the processor 901 executes the processing of the expected shopping data according to a preset personalized algorithm to obtain the user's preferred shopping data, including:

[0163] Performing expected purchase prediction processing on the expected shopping data by using the preset personalized algorithm to obtain personalized interest data of the user, wherein the personalized interest data includes the user's interest value for the current product;

[0164] Based on the user's interest value for the current commodity, the user's preferred shopping data is determined.

[0165] Optionally, the processor 901 performs the step of determining the first desired shopping data of the user based on the historical shopping behavior data and the current shopping behavior data, and the method further includes:

[0166] Performing similarity processing on the current shopping behavior and the user's historical shopping behavior;

[0167] If the similarity between the current shopping behavior and the historical shopping behavior is less than a preset similarity threshold, the current shopping behavior is used as the historical shopping behavior of the user;

[0168] If the similarity between the current shopping behavior and the historical shopping behavior is greater than or equal to a preset similarity threshold, the historical shopping behavior of the user is retained.

[0169] Optionally, the processor 901 further performs the personalized product recommendation for the user based on the preferred shopping data, and the method further includes:

[0170] Based on the current shopping behavior habits and current shopping interest data, matching associated users who have the same shopping behavior habits and shopping interest data as the user;

[0171] Acquire the associated user's preferred shopping data, wherein the preferred shopping data includes content features of preferred products and shopping behavior habits for the preferred products;

[0172] Based on the content features of the associated user's preferred products and the shopping behavior habits for the preferred products, a target product that is within a product similarity threshold interval with the preferred products is calculated, and the target product is pushed to the user.

[0173] An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the computer program implements the various processes of the personalized content push method or the application-side personalized content push method provided in the embodiment of the present invention, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0174] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0175] The above disclosure is only the preferred embodiment of the present invention, which certainly cannot be used to limit the scope of the present invention. Therefore, equivalent changes made according to the claims of the present invention are still within the scope of the present invention.

Claims

1. A method for pushing personalized content, characterized in that: include: Acquire the user's shopping behavior data and transaction behavior data, wherein the shopping behavior data includes the user's purchase history data, browsing history data, search history data, collection and preference data on the shopping page, and the user's feedback data on the product, and the transaction behavior data includes the user's payment record data, the corresponding order data of the purchased product, and the payment behavior data; Determining the user's expected shopping data based on the shopping behavior data and the transaction behavior data; Processing the expected shopping data according to a preset personalized algorithm to obtain the user's preferred shopping data, wherein the preset personalized algorithm includes a collaborative filtering algorithm, a content-based recommendation algorithm, a hybrid recommendation algorithm, and a context-aware recommendation algorithm; Based on the preferred shopping data, personalized product recommendations are made to the user; The step of processing the expected shopping data according to a preset personalized algorithm to obtain the user's preferred shopping data includes: Performing expected purchase prediction processing on the expected shopping data by using the preset personalized algorithm to obtain personalized interest data of the user, wherein the personalized interest data includes the user's interest value for the current product; Determining the user's preferred shopping data based on the user's interest value for the current product; The shopping behavior data includes historical shopping behavior data and current shopping behavior data, the transaction behavior data includes historical transaction behavior data, and the acquisition of the user's shopping behavior data and transaction behavior data includes: When a user is browsing a current product, querying the user's historical shopping behavior data for the current product, and detecting the user's current shopping behavior data when browsing the current product; Query the historical transaction behavior data of the user for the current product to obtain the transaction behavior data of the user; The determining the user's expected shopping data based on the shopping behavior data and the transaction behavior data includes: Determining first desired shopping data of the user based on the historical shopping behavior data and the current shopping behavior data; Determining second desired shopping data of the user based on the historical transaction behavior data; Determining the user's expected shopping data based on the first expected shopping data and the second expected shopping data, wherein the expected shopping data is used to determine the possibility of the user purchasing the product; The expected shopping data is obtained by analyzing the facial data of the user when browsing the current product to determine the user's emotions when browsing the current product, thereby determining whether the user has a purchase tendency for the currently browsed product. The determining of the user's first desired shopping data based on the historical shopping behavior data and the current shopping behavior data, the method further comprising: Performing similarity processing on the current shopping behavior and the user's historical shopping behavior; If the similarity between the current shopping behavior and the historical shopping behavior is less than a preset similarity threshold, the current shopping behavior is used as the historical shopping behavior of the user; If the similarity between the current shopping behavior and the historical shopping behavior is greater than or equal to a preset similarity threshold, retaining the user's historical shopping behavior; The method further comprises: providing personalized product recommendations to the user based on the preferred shopping data; Based on the current shopping behavior habits and current shopping interest data, matching associated users who have the same shopping behavior habits and shopping interest data as the user; Acquire the associated user's preferred shopping data, wherein the preferred shopping data includes content features of preferred products and shopping behavior habits for the preferred products; Based on the content features of the associated user's preferred products and the shopping behavior habits for the preferred products, a target product that is within a product similarity threshold interval with the preferred products is calculated, and the target product is pushed to the user.

2. A personalized content push device, characterized in that: include: An acquisition module is used to acquire the user's shopping behavior data and transaction behavior data, wherein the shopping behavior data includes the user's purchase history data, browsing history data, search history data, collection and preference data on the shopping page, and the user's feedback data on the product, and the transaction behavior data includes the user's payment record data, the corresponding order data of the purchased product, and the payment behavior data; A determination module, used to determine the user's expected shopping data based on the shopping behavior data and the transaction behavior data; A processing module, used to process the expected shopping data according to a preset personalized algorithm to obtain the user's preferred shopping data, wherein the preset personalized algorithm includes a collaborative filtering algorithm, a content-based recommendation algorithm, a hybrid recommendation algorithm, and a context-aware recommendation algorithm; A recommendation module, used to make personalized product recommendations to the user based on the preferred shopping data; The processing module is further used to perform expected purchase prediction processing on the expected shopping data through the preset personalized algorithm to obtain the personalized interest data of the user, wherein the personalized interest data includes the user's interest value for the current product; and determine the user's preferred shopping data based on the user's interest value for the current product; The processing module further includes: a first determination submodule, for determining the first expected shopping data of the user based on the historical shopping behavior data and the current shopping behavior data; a second determination submodule, for determining the second expected shopping data of the user based on the historical transaction behavior data; a third determination submodule, for determining the expected shopping data of the user based on the first expected shopping data and the second expected shopping data, wherein the expected shopping data is used to determine the possibility of the user purchasing the product, wherein the expected shopping data is determined by analyzing the facial data of the user when browsing the current product to determine the emotion of the user when browsing the current product, thereby determining whether the user has a purchase tendency for the currently browsed product; The acquisition module also includes a query submodule, which is used to query the user's historical shopping behavior data for the current product when the user is browsing the current product, and detect the user's current shopping behavior data when browsing the current product; the first acquisition submodule is used to query the user's historical transaction behavior data for the current product to obtain the user's transaction behavior data; The first determination submodule further includes a first processing submodule, which is used to perform similarity processing between the current shopping behavior and the historical shopping behavior of the user; a first shopping submodule, which is used to use the current shopping behavior as the historical shopping behavior of the user if the similarity between the current shopping behavior and the historical shopping behavior is less than a preset similarity threshold; and a second shopping submodule, which is used to retain the historical shopping behavior of the user if the similarity between the current shopping behavior and the historical shopping behavior is greater than or equal to the preset similarity threshold; The recommendation module also includes a first matching submodule, which is used to match associated users who have the same shopping behavior habits and shopping interest data as the user based on the current shopping behavior habits and current shopping interest data; a third acquisition submodule, which is used to acquire the preferred shopping data of the associated users, the preferred shopping data including the content features of the preferred products and the shopping behavior habits for the preferred products; and a recommendation submodule, which is used to calculate the target products that are within the product similarity threshold interval with the preferred products based on the content features of the preferred products of the associated users and the shopping behavior habits for the preferred products, and push the target products to the user.

3. A personalized content push system, characterized in that: The personalized content push includes: personalized content push device; The personalized content pushing device implements the personalized content pushing method described in claim 1.

4. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps in the method for pushing personalized content according to claim 1 when executing the computer program.

5. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the method for pushing personalized content according to claim 1 are implemented.

Images