Product recommendation method and device, computer equipment, storage medium and program product

Abstract

The invention relates to a product recommendation method and device, computer equipment, a storage medium and a program product, and relates to the technical field of artificial intelligence and big data. The method comprises the following steps: determining a plurality of candidate products according to historical consumption data of a user terminal; predicting the consumption preference of a user corresponding to the user terminal according to the related information of each candidate product to obtain a prediction result; the related information comprises product consumption place information, product type information, product consumption time and product consumption tool information; and determining a target product from the products related to the candidate products according to the prediction result, and recommending the target product to a user terminal. According to the method, one user is taken as a main body, the consumption preference of the user is predicted based on the related information of the plurality of candidate products in the historical consumption data of the user, and meanwhile, a one-to-one product recommendation service is provided for the user based on the consumption preference of the user, so that the product recommendation accuracy is greatly improved.

Description

Technical Field

[0001] The present application relates to the fields of artificial intelligence and big data technologies, and particularly to a product recommendation method, apparatus, computer device, storage medium, and program product. Background Art

[0002] With the rapid development of Internet and mobile technologies, more and more online payment methods are used by consumers, and a large amount of consumption data will be generated during the online payment process. For example, consumption merchants, payment methods, and consumption times. Providing personalized service recommendations for consumers based on their consumption data has become an important means of marketing.

[0003] In related technologies, usually, based on a huge and complex data set on the Internet, customer groups are divided into several major categories, and consumption items that they may be interested in are pushed to customers of each type according to the major categories. However, even for customers in the same major category, there are still some differences in their consumption preferences.

[0004] Therefore, the above method of recommending consumption items to users has the problem of inaccurate recommendations. Summary of the Invention

[0005] Based on this, it is necessary to provide a product recommendation method, apparatus, computer device, storage medium, and program product that can accurately recommend products for different users in view of the above technical problems.

[0006] In a first aspect, the present application provides a product recommendation method, including:

[0007] Determine a plurality of candidate products according to the historical consumption data of the user terminal;

[0008] Predict the consumption preferences of the user corresponding to the user terminal according to the relevant information of each candidate product to obtain a prediction result; the relevant information includes product consumption place information, product type information, product consumption time, and product consumption tool information;

[0009] Determine a target product from the products related to each candidate product according to the prediction result, and recommend the target product to the user terminal.

[0010] In one of the embodiments, the above relevant information includes product consumption place information and product type information, and the predicting the consumption preferences of the user corresponding to the user terminal according to the relevant information of each candidate product to obtain a prediction result includes:

[0011] Extract the product consumption place information and product type information from the relevant information of each candidate product;

[0012] Predict the consumption preferences of the user corresponding to the user terminal based on the relevant information of each candidate product, and obtain a prediction result, including:

[0013] Input the product consumption place information and product type information corresponding to each candidate product into a preset prediction network for prediction to obtain the user's high-frequency consumption place and / or high-frequency consumption type.

[0014] In one embodiment, extracting the product consumption place information and product type information from the relevant information of each candidate product includes:

[0015] Encode the relevant information of each candidate product to obtain a text representation of the relevant information of each candidate product;

[0016] Input the text representation of the relevant information of each candidate product into a preset BERT model for recognition to obtain the product consumption place information and product type information in the relevant information of each candidate product.

[0017] In one embodiment, the above relevant information includes product consumption time and product consumption place information. Predicting the consumption preferences of the user corresponding to the user terminal based on the relevant information of each candidate product, the obtained prediction result includes:

[0018] Based on time series analysis, analyze the product consumption time in the relevant information of each candidate product to determine the user's high-frequency consumption time period;

[0019] Perform a correlation analysis on the product consumption time and product consumption place information in the relevant information of each candidate product to determine the consumption place associated with the high-frequency consumption time period.

[0020] In one embodiment, the above method further includes:

[0021] Preprocess the product consumption time to obtain the preprocessed product consumption time; the preprocessing includes at least one of outlier processing, transformation processing, and redundancy processing;

[0022] Analyze the product consumption time in the relevant information of each candidate product to determine the user's high-frequency consumption time period, including:

[0023] Analyze the preprocessed product consumption time to determine the user's high-frequency consumption time period.

[0024] In one embodiment, the above relevant information includes product consumption place information. Predicting the consumption preferences of the user corresponding to the user terminal based on the relevant information of each candidate product, the obtained prediction result includes:

[0025] Analyze the consumption place relevance of the user based on the product consumption place information in the relevant information of each candidate product to obtain the consumption places associated with the user's high-frequency consumption places.

[0026] In one embodiment, the above relevant information includes product consumption tool information. Predict the consumption preferences of the user corresponding to the user terminal based on the relevant information of each candidate product to obtain a prediction result, including:

[0027] Predict the user's high-frequency consumption tools based on the product consumption tool information in the relevant information of each candidate product, and determine the type of the user's high-frequency consumption tools.

[0028] In a second aspect, the present application also provides a product recommendation device, including:

[0029] A determination module, configured to determine a plurality of candidate products according to the historical consumption data of the user terminal;

[0030] A prediction module, configured to predict the consumption preferences of the user corresponding to the user terminal based on the relevant information of each candidate product to obtain a prediction result; the relevant information includes product consumption place information, product type information, product consumption time, and product consumption tool information;

[0031] A recommendation module, configured to determine a target product from the products related to each candidate product according to the prediction result, and recommend the target product to the user terminal.

[0032] In a third aspect, the present application also provides a computer device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0033] Determine a plurality of candidate products according to the historical consumption data of the user terminal;

[0034] Predict the consumption preferences of the user corresponding to the user terminal based on the relevant information of each candidate product to obtain a prediction result; the relevant information includes product consumption place information, product type information, product consumption time, and product consumption tool information;

[0035] Determine a target product from the products related to each candidate product according to the prediction result, and recommend the target product to the user terminal.

[0036] In a fourth aspect, the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0037] Determine a plurality of candidate products according to the historical consumption data of the user terminal;

[0038] Predict the consumption preferences of the user corresponding to the user terminal based on the relevant information of each candidate product, and obtain a prediction result; the relevant information includes product consumption place information, product type information, product consumption time, and product consumption tool information;

[0039] Determine a target product from the products related to each candidate product according to the prediction result, and recommend the target product to the user terminal.

[0040] In a fifth aspect, the present application also provides a computer program product, including a computer program, which when executed by a processor, implements the following steps:

[0041] Determine a plurality of candidate products according to the historical consumption data of the user terminal;

[0042] Predict the consumption preferences of the user corresponding to the user terminal based on the relevant information of each candidate product, and obtain a prediction result; the relevant information includes product consumption place information, product type information, product consumption time, and product consumption tool information;

[0043] Determine a target product from the products related to each candidate product according to the prediction result, and recommend the target product to the user terminal.

[0044] The above product recommendation method, device, computer device, storage medium, and program product. The above method predicts the consumption preferences of the user based on the relevant information of multiple candidate products in the user's historical consumption data, and determines the target product from the candidate products related to the candidate products based on the prediction result, and recommends the target product to the user. Compared with the existing method of recommending the same consumption items that the user may be interested in for users in the same category according to the category of the customer group, the above method takes one user as the main body, predicts the consumption preferences of the user based on the relevant information of multiple candidate products in the user's historical consumption data, and at the same time provides a one-to-one product recommendation service for the user based on the user's consumption preferences, further improving the accuracy of product recommendation for the user. Description of the Drawings

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required for the description of the embodiments of the present application or related technologies. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings.

[0046] Figure 1 It is an application environment diagram of the product recommendation method in an embodiment;

[0047] Figure 2Schematic flowchart of a product recommendation method in an embodiment;

[0048] Figure 3 Schematic flowchart of a product recommendation method in another embodiment;

[0049] Figure 4 Schematic flowchart of a product recommendation method in another embodiment;

[0050] Figure 5 Schematic structural diagram of a prediction network in an embodiment;

[0051] Figure 6 Schematic flowchart of a product recommendation method in another embodiment;

[0052] Figure 7 Schematic flowchart of a product recommendation method in another embodiment;

[0053] Figure 8 Schematic flowchart of a product recommendation method in an embodiment;

[0054] Figure 9 Schematic block diagram of a product recommendation device in an embodiment. Detailed implementation manners

[0055] In order to make the objectives, technical solutions and advantages of the present application clearer and more understandable, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0056] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present application belongs; the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit the present application; the terms "including" and "having" and any variations thereof in the specification and claims of the present application and the above drawings are intended to cover non-exclusive inclusion.

[0057] In the description of the embodiments of the present application, technical terms such as "first" and "second" are only used to distinguish different objects and cannot be understood as indicating or implying relative importance or implicitly indicating the quantity, specific order or primary-secondary relationship of the indicated technical features. In the description of the embodiments of the present application, "a plurality of" means more than two unless otherwise specifically defined.

[0058] References to "embodiments" in this specification mean that a particular feature, structure, or characteristic described in connection with the embodiments can be included in at least one embodiment of the present application. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0059] With the rapid development of Internet and mobile technologies, more and more online payment methods are used by consumers. During the process of online payment, a large amount of consumption data will be generated, such as consumption merchants, payment methods, and consumption time. Providing personalized service recommendations for consumers based on their consumption data has become an important means of marketing. In related technologies, recommendation algorithms usually divide customer groups into several major categories through a large and complex data set on the Internet, and push consumption items that they may be interested in to customers according to the major categories. However, even though there are general recommendations for groups, there will still be some differences in the consumption preferences of two customers assigned to the same major category by the recommendation algorithm. Therefore, the above-mentioned recommendation algorithm has the problem of inaccuracy. The present application aims to solve this problem.

[0060] After introducing the background technology of the product recommendation method provided by the embodiments of the present application above, below, the implementation environment involved in the product recommendation method provided by the embodiments of the present application will be briefly described. The product recommendation method provided by the embodiments of the present application can be applied to, for example Figure 1In the computer device shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit, and an input device. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface, the display unit, and the input device are connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, NFC (Near Field Communication), or other technologies. The program of the computer device, when executed by the processor, implements a retrieval method. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the server can be a touch layer covered on the display screen, or a button, a trackball, or a touchpad provided on the computer device housing, or an external keyboard, touchpad, or mouse, etc.

[0061] Those skilled in the art can understand that Figure 1 the structure shown in is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the server to which the solution of this application is applied. The specific server may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0062] After introducing the application scenario of the product recommendation method provided in the embodiments of this application above, the product recommendation method described in this application will be mainly introduced below.

[0063] In one embodiment, as Figure 2 shown, a product recommendation method is provided. Taking the method applied to the Figure 1 computer device in as an example for description, it includes the following steps:

[0064] S201. Determine a plurality of candidate products according to the historical consumption data of the user terminal.

[0065] Among them, historical consumption data refers to the consumption data generated by users using online payment methods such as bank cards or payment software. The historical consumption data includes at least one of the consumption place, consumption time, consumption amount, and consumption product information. For example, the historical consumption data can be that the user uses a bank card to consume a certain amount at a certain merchant in a certain district of a certain city at a certain time and purchases a certain product.

[0066] Among them, the candidate products refer to the consumption objects in the historical consumption data. For example, if the multiple consumption objects in the historical consumption data include mobile phones, apples, and dumplings, then the multiple candidate products determined from the historical consumption data are mobile phones, apples, and dumplings.

[0067] It should be noted that the collected information is the information and data authorized by the user or fully authorized by all parties, and the processing of relevant data such as collection, storage, use, processing, transmission, provision, disclosure, and application complies with the relevant laws, regulations, and standards of relevant countries and regions. Necessary confidentiality measures are taken, which do not violate public order and good customs, and a corresponding operation entry is provided for the user to choose to authorize or refuse.

[0068] In the embodiment of the present application, every time a user generates an online consumption, the computer device can upload the consumption data generated by each online consumption to the consumption information database for storage, so as to be called at any time when analyzing the consumption data later. When it is necessary to recommend products to a certain user, the historical consumption data generated by the user can be analyzed to obtain the user's consumption preferences, and products can be recommended to the user based on the user's consumption preferences. Then, before analyzing the historical consumption data generated by the user, the computer device needs to first obtain the historical consumption data of the user, and after obtaining the historical consumption data of the user, determine multiple candidate products that the user has consumed from the historical consumption data.

[0069] Optionally, the user can directly obtain the historical consumption data of the user in the consumption information database of the computer device; for example, each user can store the historical consumption data generated by them into their respective consumption information databases. When it is necessary to obtain the historical consumption data of a certain user terminal, the user can input the user identifier of the certain user on the computer device, and then the computer device obtains the historical consumption data of the certain user according to the user identifier of the user, and then extracts multiple candidate products that the user has consumed from the historical consumption data.

[0070] S202. Predict the consumption preferences of the user corresponding to the user terminal according to the relevant information of each candidate product to obtain a prediction result.

[0071] Among them, the relevant information includes product consumption place information, product type information, product consumption time, and product consumption tool information.

[0072] In the embodiments of the present application, when obtaining multiple candidate products in the above-mentioned historical consumption data, relevant information of each candidate product can be further determined, and the consumption preferences of the user corresponding to the user terminal can be predicted based on the relevant information of each candidate product to obtain a prediction result of the user's consumption preferences; it should be noted that the relevant information of each candidate product can be input into a trained prediction model for prediction to obtain the user's consumption preference information. Optionally, the high-frequency consumption time period of the user corresponding to the user terminal can be predicted based on the consumption occurrence time of each candidate product to obtain the high-frequency consumption time period of the user of the user terminal; optionally, the high-frequency consumption occurrence place of the user corresponding to the user terminal can also be predicted based on the consumption occurrence place information of each candidate product to obtain the high-frequency consumption place of the user of the user terminal; optionally, the high-frequency consumption product of the user corresponding to the user terminal can also be predicted based on the consumption product information of each candidate product to obtain the high-frequency consumption product of the user of the user terminal.

[0073] S203. Determine a target product from the products related to each candidate product according to the prediction result, and recommend the target product to the user terminal.

[0074] Among them, the products related to each candidate product refer to products of the same product type as each candidate product. For example, if the candidate product is a mobile phone, the products related to the candidate product can be electronic products such as headphones, power banks, tablet computers, chargers, and laptop computers; for example, if the candidate product is an apple, the products related to the candidate product can be fruits such as bananas, honeydews, grapes, and tomatoes.

[0075] In the embodiments of the present application, after determining the prediction result of the user's consumption preferences as described above, a target product can be screened from the products related to each candidate product according to the prediction result of the user's consumption preferences, and the target product can be recommended to the user terminal. For example, if the user's consumption preference is for electronic products, electronic products can be screened from the products related to each candidate product as the target product and recommended to the user.

[0076] The product recommendation method provided by the embodiments of the present application determines multiple candidate products according to the historical consumption data of the user terminal; predicts the consumption preferences of the user corresponding to the user terminal according to the relevant information of each candidate product to obtain a prediction result; the relevant information includes product consumption place information, product type information, product consumption time, and product consumption tool information; determines a target product from the products related to each candidate product according to the prediction result, and recommends the target product to the user terminal. The above method predicts the consumption preferences of the user based on the relevant information of multiple candidate products in the user's historical consumption data, and determines the target product from the candidate products related to the candidate products based on the prediction result, and recommends the target product to the user. Compared with the existing method of recommending the same consumption items that the user may be interested in according to the category of the customer group, the above method takes one user as the main body, predicts the consumption preferences of the user based on the relevant information of multiple candidate products in the historical consumption data of the user, and at the same time provides a one-to-one product recommendation service for the user based on the consumption preferences of the user, further improving the accuracy of product recommendation for the user.

[0077] In one embodiment, on the basis of the embodiment shown in Figure 2 The above relevant information includes product consumption place information and product type information, and the determination process of the prediction result can be described. For example, Figure 3 As shown, the above S202 "predict the consumption preferences of the user corresponding to the user terminal according to the relevant information of each candidate product to obtain a prediction result" includes:

[0078] S301. Extract product consumption place information and product type information from the relevant information of each candidate product.

[0079] In the embodiments of the present application, after obtaining multiple candidate products in the historical consumption data, the relevant information of each candidate product is determined, and then the product consumption place information and product type information are extracted from the relevant information of each candidate product. Optionally, the product consumption place information and product type information corresponding to the candidate product can be queried from the relevant information of the candidate product by means of manual query.

[0080] Optionally, the following provides a method for obtaining product consumption place information and product type information, that is, S301 "extract product consumption place information and product type information from the relevant information of each candidate product". As shown in Figure 4 It includes:

[0081] S401. Encode the relevant information of each candidate product to obtain a text representation of the relevant information of each candidate product.

[0082] In the embodiments of the present application, after obtaining the relevant information of each candidate product, the Bidirectional Encoder Representations from Transformers (BERT) model can be used to encode the relevant information of each candidate product to obtain the text representation of the relevant information of each candidate product.

[0083] S402. Input the text representation of the relevant information of each candidate product into a preset BERT model for recognition to obtain the product consumption place information and product type information in the relevant information of each candidate product.

[0084] In the embodiments of the present application, the BERT model can be pre-trained so that the trained preset BERT model has the function of recognizing the relevant information of the input candidate product to obtain the product consumption place information and product type information in the relevant information of the candidate product. Then, after the preset BERT model is trained, the text representation of the relevant information of the candidate product can be input into the preset BERT model for recognition to obtain the product consumption place information and product type information in the relevant information of each candidate product.

[0085] The above "predicting the consumption preference of the user corresponding to the user terminal according to the relevant information of each candidate product to obtain a prediction result" includes:

[0086] S302. Input the product consumption place information and product type information corresponding to each candidate product into a preset prediction network for prediction to obtain the high-frequency consumption place and / or high-frequency consumption type of the user.

[0087] Among them, the preset prediction network can be constructed according to the Long Short-Term Memory (LSTM) and the neural network ensemble attention mechanism. The reason for choosing LSTM is that LSTM can better capture sequence information and process long sequences compared with other algorithms, and the reason for choosing the neural network ensemble attention mechanism is that it can better achieve personalized prediction and improve the interpretability of the model. In the embodiments of the present application, combining LSTM with the neural network ensemble attention mechanism can better capture the key information in the input sequence, reduce the interference of noise and interference on the model, and improve the generalization ability of the model on different data sets. For example, the architecture of the preset prediction network is as Figure 5 shown, including a BERT word embedding layer, an LSTM layer, an attention layer, and an output layer.

[0088] It should be noted that the prediction network can be pre-trained so that the trained preset prediction network has the function of inputting product consumption place information and product type information into the preset prediction network for prediction to obtain the user's high-frequency consumption place and / or high-frequency consumption type.

[0089] Among them, the high-frequency consumption place refers to the place where the probability that the user arrives and consumes is greater than the first preset probability value, and the high-frequency consumption type refers to the type of product consumption where the probability that the user generates product consumption is greater than the second preset probability value. It should be noted that the determination process of the first preset value and the second preset value is not limited here, as long as the high-frequency consumption place and the high-frequency consumption type can be distinguished.

[0090] In the embodiment of the present application, after obtaining the product consumption place information and product type information corresponding to each candidate product, the product consumption place information and product type information corresponding to each candidate product can be input into the preset prediction network for prediction to obtain the user's high-frequency consumption place and / or high-frequency consumption type.

[0091] The prediction method for high-frequency consumption place and / or high-frequency consumption type provided by the embodiment of the present application predicts the high-frequency consumption place and / or high-frequency consumption type based on the relevant information of candidate products in historical consumption data and a preset prediction model to obtain a prediction result, providing a basis for subsequently recommending target products to the user based on the prediction result.

[0092] In one embodiment, on the basis of Figure 2 the embodiment shown, the above-mentioned relevant information includes product consumption time and product consumption place information, and the determination process of the prediction result can be described. For example, Figure 6 as shown, the above-mentioned S202 "predict the consumption preference of the user corresponding to the user terminal according to the relevant information of each candidate product to obtain a prediction result" includes:

[0093] S501. Based on time series analysis, analyze the product consumption time in the relevant information of each candidate product to determine the user's high-frequency consumption time period.

[0094] Among them, the high-frequency consumption time period refers to the time period when the probability that the user consumes is greater than the third preset probability value. It should be noted that the determination process of the third preset value is not limited here, as long as the high-frequency consumption time period can be distinguished.

[0095] In the embodiment of the present application, after obtaining multiple candidate products in the historical consumption data, the relevant information of each candidate product is determined, and then the product consumption time is extracted from the relevant information of each candidate product, and the product consumption time is analyzed based on the time series analysis method to obtain the high-frequency consumption time period of the user. For example, the high-frequency consumption time period of the user is from 18:00 to 20:00 every Saturday afternoon.

[0096] S502. Perform a correlation analysis on the product consumption time and the product consumption place information in the relevant information of each candidate product to determine the consumption place associated with the high-frequency consumption time period.

[0097] In the embodiment of the present application, after obtaining multiple candidate products in the historical consumption data, the relevant information of each candidate product is determined, and then the product consumption time and the product consumption place information are extracted from the relevant information of each candidate product, and a correlation analysis is performed on the extracted product consumption time and the product consumption place information to determine the consumption place associated with the high-frequency consumption time period. For example, the user has frequent cinema consumption records from 18:00 to 20:00 every Friday.

[0098] The prediction method for high-frequency consumption time and / or high-frequency consumption place provided by the embodiment of the present application predicts the high-frequency consumption time and the consumption place associated with the high-frequency consumption time based on the relevant information of the candidate products in the historical consumption data and the time series analysis method to obtain a prediction result, providing a basis for subsequently recommending target products to the user based on the prediction result.

[0099] In one embodiment, on the basis of Figure 6 the embodiment shown, as Figure 7 shown, the above method further includes:

[0100] S503. Preprocess the product consumption time to obtain the preprocessed product consumption time.

[0101] Among them, the preprocessing includes at least one of outlier processing, transformation processing, and redundancy processing. The preprocessing also includes the extraction and aggregation of timestamps, and the processing method of detecting and processing outliers to exclude the influence of abnormal situations on the analysis result.

[0102] In the embodiment of the present application, after obtaining multiple candidate products in the historical consumption data, the relevant information of each candidate product is determined, and then the product consumption time is extracted from the relevant information of each candidate product, and the product consumption time is preprocessed to obtain the preprocessed product consumption time.

[0103] The above S501 "Analyze the product consumption time in the relevant information of each candidate product to determine the high-frequency consumption time period of the user" includes:

[0104] S501. Analyze the product consumption time after preprocessing to determine the high-frequency consumption time period of the user.

[0105] In the embodiment of the present application, after obtaining the preprocessed product consumption time as described above, analyze the preprocessed product consumption time based on the time series analysis method to obtain the high-frequency consumption time period of the user. For example, the high-frequency consumption time period of the user is from 18:00 to 20:00 on Saturday afternoons every week.

[0106] The method for preprocessing the product consumption time provided by the embodiment of the present application can perform data cleaning, redundancy processing, etc. on the product consumption time, providing a data basis for accurately determining the high-frequency consumption time of the user and the consumption place associated with the high-frequency consumption time based on the preprocessed product consumption time.

[0107] In one embodiment, on the basis of Figure 2 the embodiment shown, the above-mentioned relevant information includes product consumption place information, and the determination process of the prediction result can be described. The above S202 "Predict the consumption preference of the user corresponding to the user terminal according to the relevant information of each candidate product to obtain a prediction result" includes:

[0108] Analyze the consumption place relevance of the user according to the product consumption place information in the relevant information of each candidate product to obtain the consumption place associated with the high-frequency consumption place of the user.

[0109] In the embodiment of the present application, after obtaining multiple candidate products in the historical consumption data as described above, determine the relevant information of each candidate product, then extract the product consumption place information corresponding to each candidate product from the relevant information of each candidate product, and analyze the consumption place relevance of the user according to the product consumption place information corresponding to each candidate product to obtain the high-frequency consumption place of the user and the consumption place associated with the high-frequency consumption place of the user. Optionally, perform place relevance analysis on the product consumption place information corresponding to each candidate product respectively, mine the association relationship between the product consumption place information corresponding to each candidate product through association rules, and analyze the association relationship to obtain the high-frequency consumption place of the user and the consumption place associated with the high-frequency consumption place of the user. For example, it is determined that the high-frequency consumption place of the user is area A, and then based on the association relationship of the consumption place, area B with a relatively large relevance to area A is determined. For example, the user often goes to area B to watch a movie after having a meal in area A.

[0110] The prediction method for the consumption place associated with the high-frequency consumption place of the user provided by the embodiment of the present application analyzes the consumption place relevance of each candidate product based on the relevance of the consumption place to obtain a prediction result, providing a basis for recommending target products to the user based on the prediction result subsequently.

[0111] In one embodiment, based on Figure 2 the embodiment shown, the above relevant information includes product consumption tool information, which can describe the determination process of the prediction result. The above S202, "Predict the consumption preference of the user corresponding to the user terminal according to the relevant information of each candidate product to obtain a prediction result", includes:

[0112] Predict the high-frequency consumption tool of the user according to the product consumption tool information in the relevant information of each candidate product, and determine the type of the high-frequency consumption tool of the user.

[0113] Among them, the product consumption tool information refers to the payment tool information used by the user when consuming the product. For example, information such as bank cards and payment software.

[0114] In the embodiment of the present application, after obtaining multiple candidate products in the historical consumption data, determine the relevant information of each candidate product, then extract the product consumption tool information corresponding to each candidate product from the relevant information of each candidate product, and analyze the high-frequency consumption tool of the user according to the product consumption tool information corresponding to each candidate product to obtain the type of the high-frequency consumption tool of the user. Optionally, the product consumption tool information in the relevant information of each candidate product can be input into a trained prediction model for prediction to obtain the high-frequency consumption tool information of the user. For example, if the product consumption tool information in the relevant information of each candidate product is input into a trained prediction model for prediction, and the high-frequency consumption tool information of the user is obtained as the payment information of a certain payment software, then when recommending products to the user, products of merchants that can be paid using this payment software can be recommended, providing the possibility for the user to purchase this product.

[0115] The method for predicting the high-frequency consumption tool information of the user provided by the embodiment of the present application provides the possibility for the subsequent recommendation of target products that can be paid based on the prediction result.

[0116] In one embodiment, a method for recommending a target product to a user terminal is further provided, including:

[0117] Step 1. After obtaining the high-frequency consumption place and / or high-frequency consumption type of the user through a preset prediction network, when carrying out a marketing activity, the high-frequency consumption place and / or high-frequency consumption type can be selected to carry out the marketing activity on the user;

[0118] Step 2. Based on the consumption place associated with the high-frequency consumption place of the user, when carrying out an activity on the high-frequency consumption place, also consider the consumption place associated with the high-frequency consumption place of the user, and conduct marketing on the consumption place associated with the high-frequency consumption place of the user;

[0119] Step 3: When carrying out marketing activities based on the user's high-frequency consumption time period and the consumption places associated with the high-frequency consumption time period, the high-frequency consumption time period and the consumption places associated with the high-frequency consumption time period can be selected to carry out marketing activities for the user. For example, if consumer A frequently uses Meituan to order takeout between 17:00 and 19:00 every day, then marketing activities for food and takeout can be carried out 1-2 hours before 17:00, making it more likely for them to accept the marketing.

[0120] In one embodiment, a complete product recommendation method is also provided, as Figure 8 shown, including:

[0121] (1) Obtain the user's historical consumption data, which includes information such as consumption places, payment methods, and consumption times;

[0122] (2) For the consumption time information, based on the time series algorithm and the user's consumption time pattern, determine the user's high-frequency consumption time period;

[0123] (3) Extract the consumption places and payment methods in the historical consumption data based on the Bert model;

[0124] (4) Based on a preset prediction network, consumption places, and payment methods, predict the user's high-frequency consumption places and / or high-frequency consumption types;

[0125] (5) Analyze the relevance of the user's consumption places according to the product consumption place information in the relevant information of each candidate product to obtain the consumption places associated with the user's high-frequency consumption places;

[0126] (6) Recommend target products for the user according to the user's high-frequency consumption time period, high-frequency consumption places, high-frequency consumption types, and the consumption places associated with the user's high-frequency consumption places.

[0127] It should be understood that although the various steps in the flowcharts involved in the above-described embodiments are shown in sequence according to the arrows, these steps do not necessarily have to be executed in the order indicated by the arrows. Unless specifically stated in this article, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, at least some of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages do not necessarily have to be executed at the same time, but can be executed at different times. The execution order of these steps or stages does not necessarily have to be sequential, but can be executed alternately or in turn with at least a part of other steps or steps or stages in other steps.

[0128] Based on the same inventive concept, an embodiment of the present application further provides a product recommendation device for implementing the product recommendation method involved above. The solution provided by this device for solving problems is similar to the solution described in the above method. Therefore, the specific limitations in one or more embodiments of the product recommendation device provided below can refer to the limitations on the product recommendation method in the above text and will not be repeated here.

[0129] In an exemplary embodiment, as Figure 9 shown, a product recommendation device is provided, including: a determination module 10, a prediction module 11, and a recommendation module 12, where:

[0130] The determination module 10 is configured to determine a plurality of candidate products according to the historical consumption data of the user terminal;

[0131] The prediction module 11 is configured to predict the consumption preference of the user corresponding to the user terminal according to the relevant information of each candidate product, and obtain a prediction result; the relevant information includes product consumption place information, product type information, product consumption time, and product consumption tool information.

[0132] The recommendation module 12 is configured to determine a target product from the products related to each candidate product according to the prediction result, and recommend the target product to the user terminal.

[0133] In an exemplary embodiment, the above relevant information includes product consumption place information and product type information. The above prediction module includes: an extraction unit, specifically configured to extract product consumption place information and product type information from the relevant information of each candidate product;

[0134] The above prediction module is further configured to input the product consumption place information and product type information corresponding to each candidate product into a preset prediction network for prediction, and obtain the high-frequency consumption place and / or high-frequency consumption type of the user.

[0135] In an exemplary embodiment, the above extraction unit is specifically further configured to encode the relevant information of each candidate product to obtain a text representation of the relevant information of each candidate product; input the text representation of the relevant information of each candidate product into a preset BERT model for recognition, and obtain the product consumption place information and product type information in the relevant information of each candidate product.

[0136] In an exemplary embodiment, the above relevant information includes product consumption time and product consumption place information. The above prediction module further includes: an analysis unit and a determination unit, where:

[0137] The analysis unit is specifically configured to analyze the product consumption time in the relevant information of each candidate product based on time series analysis, and determine the high-frequency consumption time period of the user;

[0138] A determination unit, specifically configured to perform a correlation analysis on the product consumption time and the product consumption place information in the relevant information of each candidate product, and determine the consumption place associated with the high-frequency consumption time period.

[0139] In an exemplary embodiment, the above prediction module further includes: a preprocessing unit, specifically configured to preprocess the product consumption time to obtain the preprocessed product consumption time; the preprocessing includes at least one of outlier processing, transformation processing, and redundancy processing;

[0140] The above analysis unit is further specifically configured to analyze the preprocessed product consumption time to determine the high-frequency consumption time period of the user.

[0141] In an exemplary embodiment, the above relevant information includes product consumption place information, and the above prediction module is further configured to analyze the consumption place correlation of the user according to the product consumption place information in the relevant information of each candidate product, and obtain the consumption place associated with the high-frequency consumption place of the user.

[0142] In an exemplary embodiment, the above relevant information includes product consumption tool information, and the above prediction module is further configured to predict the high-frequency consumption tool of the user according to the product consumption tool information in the relevant information of each candidate product, and determine the type of the high-frequency consumption tool of the user.

[0143] Each module in the above product recommendation device can be implemented in whole or in part by software, hardware, and their combination. The above modules can be embedded in the processor in the computer device in the form of hardware or be independent of it, or can be stored in the memory in the computer device in the form of software, so as to be called by the processor to execute the operations corresponding to the above respective modules.

[0144] In an exemplary embodiment, a computer device is provided. The computer device can be a terminal, and its internal structure diagram can be as Figure 1As shown in the figure. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit, and an input device. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface, the display unit, and the input device are connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, near field communication (NFC), or other technologies. When the computer program is executed by the processor, it implements a product recommendation method. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer covered on the display screen, or a button, a trackball, or a touchpad provided on the computer device housing, or an external keyboard, touchpad, or mouse, etc.

[0145] Those skilled in the art can understand that Figure 1 the structure shown in the figure is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0146] In an exemplary embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory. When the processor executes the computer program, the following steps are implemented:

[0147] Determine a plurality of candidate products according to the historical consumption data of the user terminal;

[0148] Predict the consumption preferences of the user corresponding to the user terminal according to the relevant information of each candidate product to obtain a prediction result; the relevant information includes product consumption place information, product type information, product consumption time, and product consumption tool information;

[0149] Determine a target product from the products related to each candidate product according to the prediction result, and recommend the target product to the user terminal.

[0150] In one embodiment, when the processor executes the computer program, the following steps are also implemented:

[0151] Extract product consumption venue information and product type information from the relevant information of each candidate product;

[0152] Predict the consumption preferences of the user corresponding to the user terminal based on the relevant information of each candidate product, and obtain a prediction result, including:

[0153] Input the product consumption venue information and product type information corresponding to each candidate product into a preset prediction network for prediction, and obtain the high-frequency consumption venues and / or high-frequency consumption types of the user.

[0154] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0155] Encode the relevant information of each candidate product to obtain a text representation of the relevant information of each candidate product;

[0156] Input the text representation of the relevant information of each candidate product into a preset BERT model for recognition, and obtain the product consumption venue information and product type information in the relevant information of each candidate product.

[0157] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0158] Based on time series analysis, analyze the product consumption time in the relevant information of each candidate product to determine the high-frequency consumption time period of the user;

[0159] Perform a correlation analysis on the product consumption time and the product consumption venue information in the relevant information of each candidate product to determine the consumption venues associated with the high-frequency consumption time period.

[0160] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0161] Preprocess the product consumption time to obtain the preprocessed product consumption time; the preprocessing includes at least one of outlier processing, transformation processing, and redundancy processing;

[0162] Analyze the product consumption time in the relevant information of each candidate product to determine the high-frequency consumption time period of the user, including:

[0163] Analyze the preprocessed product consumption time to determine the high-frequency consumption time period of the user.

[0164] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0165] Analyze the consumption place relevance of the user based on the product consumption place information in the relevant information of each candidate product to obtain the consumption places associated with the user's high-frequency consumption places.

[0166] In one embodiment, when the processor executes the computer program, the following steps are also implemented:

[0167] Predict the user's high-frequency consumption tools based on the product consumption tool information in the relevant information of each candidate product, and determine the types of the user's high-frequency consumption tools.

[0168] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0169] Determine multiple candidate products according to the historical consumption data of the user terminal;

[0170] Predict the consumption preferences of the user corresponding to the user terminal based on the relevant information of each candidate product to obtain a prediction result; the relevant information includes product consumption place information, product type information, product consumption time, and product consumption tool information;

[0171] Determine a target product from the products related to each candidate product according to the prediction result, and recommend the target product to the user terminal.

[0172] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0173] Extract the product consumption place information and product type information from the relevant information of each candidate product;

[0174] Predict the consumption preferences of the user corresponding to the user terminal based on the relevant information of each candidate product to obtain a prediction result, including:

[0175] Input the product consumption place information and product type information corresponding to each candidate product into a preset prediction network for prediction to obtain the user's high-frequency consumption places and / or high-frequency consumption types.

[0176] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0177] Encode the relevant information of each candidate product to obtain a text representation of the relevant information of each candidate product;

[0178] Input the text representation of the relevant information of each candidate product into a preset BERT model for recognition to obtain the product consumption place information and product type information in the relevant information of each candidate product.

[0179] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0180] Based on time series analysis, analyze the product consumption time in the relevant information of each candidate product to determine the user's high-frequency consumption time period;

[0181] Perform a correlation analysis on the product consumption time and the product consumption place information in the relevant information of each candidate product to determine the consumption places associated with the high-frequency consumption time period.

[0182] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0183] Preprocess the product consumption time to obtain the preprocessed product consumption time; the preprocessing includes at least one of outlier processing, transformation processing, and redundancy processing;

[0184] Analyze the product consumption time in the relevant information of each candidate product to determine the user's high-frequency consumption time period, including:

[0185] Analyze the preprocessed product consumption time to determine the user's high-frequency consumption time period.

[0186] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0187] Analyze the user's consumption place correlation based on the product consumption place information in the relevant information of each candidate product to obtain the consumption places associated with the user's high-frequency consumption places.

[0188] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0189] Predict the user's high-frequency consumption tools based on the product consumption tool information in the relevant information of each candidate product to determine the types of the user's high-frequency consumption tools.

[0190] In one embodiment, a computer program product is provided, including a computer program, which when executed by a processor, implements the following steps:

[0191] Determine multiple candidate products based on the historical consumption data of the user terminal;

[0192] Predict the user's consumption preferences corresponding to the user terminal based on the relevant information of each candidate product to obtain a prediction result; the relevant information includes product consumption place information, product type information, product consumption time, and product consumption tool information;

[0193] Determine the target product from the products related to each candidate product according to the prediction result, and recommend the target product to the user terminal.

[0194] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0195] Extract product consumption place information and product type information from the relevant information of each candidate product;

[0196] Predict the consumption preferences of the user corresponding to the user terminal according to the relevant information of each candidate product, and obtain a prediction result, including:

[0197] Input the product consumption place information and product type information corresponding to each candidate product into a preset prediction network for prediction, and obtain the user's high-frequency consumption places and / or high-frequency consumption types.

[0198] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0199] Encode the relevant information of each candidate product to obtain a text representation of the relevant information of each candidate product;

[0200] Input the text representation of the relevant information of each candidate product into a preset BERT model for recognition, and obtain the product consumption place information and product type information in the relevant information of each candidate product.

[0201] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0202] Based on time series analysis, analyze the product consumption time in the relevant information of each candidate product to determine the user's high-frequency consumption time period;

[0203] Conduct a correlation analysis on the product consumption time and product consumption place information in the relevant information of each candidate product to determine the consumption places associated with the high-frequency consumption time period.

[0204] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0205] Preprocess the product consumption time to obtain the preprocessed product consumption time; the preprocessing includes at least one of outlier processing, transformation processing, and redundancy processing;

[0206] Analyze the product consumption time in the relevant information of each candidate product to determine the user's high-frequency consumption time period, including:

[0207] Analyze the preprocessed product consumption time to determine the user's high-frequency consumption time period.

[0208] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0209] Analyze the consumption place relevance of the user according to the product consumption place information in the relevant information of each candidate product, and obtain the consumption places associated with the user's high-frequency consumption places.

[0210] In one embodiment, when the processor executes the computer program, the following steps are also implemented:

[0211] Predict the user's high-frequency consumption tools according to the product consumption tool information in the relevant information of each candidate product, and determine the types of the user's high-frequency consumption tools.

[0212] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with relevant regulations.

[0213] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, data processing logics based on quantum computing, artificial intelligence (AI) processors, etc., without limitation.

[0214] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered to be within the scope recorded in the present application.

[0215] The above-described embodiments merely represent several implementation manners of the present application. The description thereof is relatively specific and detailed, but it should not be construed as a limitation on the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all fall within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.

Claims

1. A product recommendation method, characterized in that, The method includes: Determining a plurality of candidate products based on the historical consumption data of the user terminal; Predicting the consumption preferences of the user corresponding to the user terminal according to the relevant information of each candidate product to obtain a prediction result; the relevant information includes product consumption place information, product type information, product consumption time, and product consumption tool information; Determining a target product from the products related to each candidate product according to the prediction result, and recommending the target product to the user terminal.

2. The method according to claim 1, characterized in that, The relevant information includes product consumption place information and product type information. The predicting the consumption preferences of the user corresponding to the user terminal according to the relevant information of each candidate product to obtain a prediction result includes: Extracting product consumption place information and product type information from the relevant information of each candidate product; The predicting the consumption preferences of the user corresponding to the user terminal according to the relevant information of each candidate product to obtain a prediction result includes: Inputting the product consumption place information and product type information corresponding to each candidate product into a preset prediction network for prediction to obtain the high-frequency consumption places and / or high-frequency consumption types of the user.

3. The method according to claim 2, wherein The extracting product consumption place information and product type information from the relevant information of each candidate product includes: Encoding the relevant information of each candidate product to obtain a text representation of the relevant information of each candidate product; Inputting the text representation of the relevant information of each candidate product into a preset BERT model for recognition to obtain the product consumption place information and product type information in the relevant information of each candidate product.

4. The method according to claim 1, wherein The relevant information includes product consumption time and product consumption place information. The predicting the consumption preferences of the user corresponding to the user terminal according to the relevant information of each candidate product to obtain a prediction result includes: Analyzing the product consumption time in the relevant information of each candidate product based on time series analysis to determine the high-frequency consumption time period of the user; Performing a correlation analysis on the product consumption time and product consumption place information in the relevant information of each candidate product to determine the consumption places associated with the high-frequency consumption time period.

5. The method according to claim 4, wherein The method further includes: Performing preprocessing on the product consumption time to obtain preprocessed product consumption time; the preprocessing includes at least one of outlier processing, transformation processing, and redundancy processing; The analyzing the product consumption time in the relevant information of each candidate product to determine the high-frequency consumption time period of the user includes: Analyzing the preprocessed product consumption time to determine the high-frequency consumption time period of the user.

6. The method according to claim 1, wherein The relevant information includes product consumption place information. The predicting the consumption preferences of the user corresponding to the user terminal according to the relevant information of each candidate product to obtain a prediction result includes: Analyzing the consumption place correlation of the user according to the product consumption place information in the relevant information of each candidate product to obtain the consumption places associated with the high-frequency consumption places of the user.

7. The method according to claim 1, characterized in that, The relevant information includes product consumption tool information. Predicting the consumption preferences of the user corresponding to the user terminal according to the relevant information of each candidate product to obtain a prediction result includes: Predicting the high-frequency consumption tool of the user according to the product consumption tool information in the relevant information of each candidate product, and determining the type of the high-frequency consumption tool of the user.

8. A product recommendation device, characterized in that, The device includes: A determination module, configured to determine a plurality of candidate products according to the historical consumption data of the user terminal; A prediction module, configured to predict the consumption preferences of the user corresponding to the user terminal according to the relevant information of each candidate product to obtain a prediction result; the relevant information includes product consumption place information, product type information, product consumption time, and product consumption tool information; A recommendation module, configured to determine a target product from the products related to each candidate product according to the prediction result, and recommend the target product to the user terminal.

9. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

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

Images