Method, apparatus, storage medium, and electronic device for determining commodity purchase intention

Abstract

The present disclosure relates to a method, apparatus, storage medium, and electronic device for determining the intention to purchase a commodity, and relates to the field of intelligent learning technology. The method includes: determining user characteristic information of a user to be screened for a preset type of commodity; based on the user characteristic information, determining a target user among the users to be screened who has a target purchase intention, where the target purchase intention refers to the intention of purchasing a target commodity among the preset type of commodities. The beneficial effects of the present disclosure are: through multi-dimensional user characteristic information, it is possible to accurately predict the magnitude of the intention of the user to be screened to purchase the target commodity among the preset type of commodities, so as to accurately locate potential target users among the users to be screened, provide data support for subsequent marketing strategies, and improve the success rate of marketing strategies.

Description

Technical Field

[0001] The present disclosure relates to the field of intelligent learning technologies, and in particular, to a method, apparatus, storage medium, and electronic device for determining a commodity purchase intention. Background Art

[0002] In the current marketing process, it is generally necessary to collect a large amount of user data. By classifying the user data, user portraits corresponding to user groups can be obtained. When a marketing activity needs to be carried out, marketing strategies are sent according to the user portraits. However, this marketing method does not consider whether a single user has the intention to purchase a commodity, let alone whether a single user has the intention to purchase a commodity of a specific brand, resulting in low marketing efficiency. Summary of the Invention

[0003] To overcome the problems existing in the related art, the present disclosure provides a method, apparatus, storage medium, and electronic device for determining a commodity purchase intention.

[0004] According to a first aspect of an embodiment of the present disclosure, there is provided a method for determining a commodity purchase intention, including:

[0005] Determining user feature information of a user to be screened for a preset type of commodity;

[0006] Based on the user feature information, determining target users among the users to be screened who have a target purchase intention, where the target purchase intention refers to the intention of purchasing a target commodity among the preset type of commodities.

[0007] In some embodiments, the user feature information includes at least one of the following:

[0008] User portrait information, behavior feature information, and advertisement push feature information, where the behavior feature information represents operation statistical information of the user for advertisement messages related to the preset type of commodity, and the advertisement push feature information represents attribute information of advertisement messages related to the preset type of commodity received by the user.

[0009] In some embodiments, the user feature information includes user portrait information, behavior feature information, and advertisement push feature information;

[0010] The determining, based on the user feature information, target users among the users to be screened who have a target purchase intention includes:

[0011] Inputting the user portrait information, the behavior feature information, and the advertisement push feature information into a trained commodity purchase prediction model to determine target users among the users to be screened who have a target purchase intention.

[0012] In some embodiments, the commodity purchase prediction model includes a first sub-model and a second sub-model, wherein the first sub-model is used to extract a feature vector characterizing the intention of the user to be screened to purchase the commodity of the preset type, and the second sub-model is used to extract a feature vector characterizing the intention of the user to be screened to purchase the target commodity;

[0013] Inputting the user portrait information, the behavioral feature information, and the advertisement push feature information into the trained commodity purchase prediction model to determine the target users among the users to be screened who have the target purchase intention includes:

[0014] Encoding the user portrait information, the behavioral feature information, and the advertisement push feature information respectively to obtain a user feature vector, a behavioral feature vector, and an advertisement push feature vector;

[0015] Concatenating the user portrait vector, the behavioral feature vector, and the advertisement push feature vector to obtain a comprehensive feature vector;

[0016] Taking the comprehensive feature vector as the input of the first sub-model and the second sub-model respectively, and extracting a first feature vector and a second feature vector;

[0017] Based on the first feature vector and the second feature vector, determining the target users among the users to be screened who have the target purchase intention.

[0018] In some embodiments, the method further includes:

[0019] Concatenating the first feature vector and the second feature vector to obtain a concatenated vector;

[0020] The determining the target users among the users to be screened who have the target purchase intention based on the first feature vector and the second feature vector includes:

[0021] Based on the first feature vector, the second feature vector, and the concatenated vector, determining the target users among the users to be screened who have the target purchase intention.

[0022] In some embodiments, the determining the target users among the users to be screened who have the target purchase intention based on the first feature vector, the second feature vector, and the concatenated vector includes:

[0023] Based on the first feature vector, determining a first score, where the first score characterizes the magnitude of the intention of the user to be screened to purchase the commodity of the preset type;

[0024] Based on the second feature vector, determine a second score, where the second score represents the magnitude of the intention of the user to be screened to purchase the target product.

[0025] Based on the concatenated vector, determine a third score, where the third score represents the magnitude of the intention of the user to be screened to purchase the target product among the products of the preset type and to purchase the products of the preset type.

[0026] Based on the maximum value among the first score, the second score, and the third score, determine the target users among the users to be screened who have the target purchase intention.

[0027] In some embodiments, the first sub-model and the second sub-model are CNN convolutional neural network models.

[0028] According to a second aspect of the embodiments of the present disclosure, there is provided a device for determining the intention of purchasing a product, including:

[0029] An acquisition module configured to determine user feature information of a user to be screened for products of a preset type.

[0030] An intention determination module configured to determine, based on the user feature information, the target users among the users to be screened who have the target purchase intention, where the target purchase intention refers to the intention of purchasing the target product among the products of the preset type.

[0031] According to a third aspect of the embodiments of the present disclosure, there is provided a computer-readable storage medium, on which computer program instructions are stored, and when the program instructions are executed by a processor, the steps of the method for determining the intention of purchasing a product provided in the first aspect of the present disclosure are implemented.

[0032] According to a fourth aspect of the embodiments of the present disclosure, there is provided an electronic device, including:

[0033] A memory on which a computer program is stored.

[0034] A processor for executing the computer program in the memory to implement the steps of the method for determining the intention of purchasing a product provided in the first aspect of the present disclosure.

[0035] The technical solutions provided by the embodiments of the present disclosure may include the following beneficial effects: Through multi-dimensional user feature information, it is possible to accurately predict the magnitude of the intention of the user to be screened to purchase the target product among the products of the preset type, so as to accurately locate potential target users among the users to be screened, provide data support for subsequent marketing strategies, and improve the success rate of marketing strategies. For example, after locating potential target users, it is possible to prompt target users to purchase the target product by accurately delivering coupons, pushing advertising, etc.

[0036] It should be understood that the above general description and the following detailed description are merely exemplary and explanatory, and do not limit the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The accompanying drawings herein are incorporated into and constitute a part of this specification, showing embodiments consistent with the present disclosure, and together with the specification are used to explain the principles of the present disclosure.

[0038] Figure 1 is a flowchart of a method for determining a commodity purchase intention shown according to an exemplary embodiment.

[0039] Figure 2 is a flowchart of determining a target user based on a commodity purchase prediction model shown according to an exemplary embodiment.

[0040] Figure 3 is an architecture diagram of a commodity purchase prediction model shown according to an exemplary embodiment.

[0041] Figure 4 is a block diagram of a device for determining a commodity purchase intention shown according to an exemplary embodiment.

[0042] Figure 5 is a block diagram of an electronic device shown according to an exemplary embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0043] Here, the exemplary embodiments will be described in detail, and the examples are shown in the accompanying drawings. When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0044] Figure 1 is a flowchart of a method for determining a commodity purchase intention shown according to an exemplary embodiment. As Figure 1 shown, this method for determining a commodity purchase intention is used in a terminal or a server and includes the following steps.

[0045] In step 110, determine the user characteristic information of the user to be screened for commodities of a preset type.

[0046] Here, the users to be filtered can refer to the user group of a certain type of commodity. For example, the users to be filtered can be the user group of mobile phones. Of course, the users to be filtered can also be the user group of a specific merchant. For example, the users of Xiaomi mobile phones. Among them, the users to be filtered can include at least one user. The preset type of commodity refers to a specific category of commodity. For example, the preset type of commodity can refer to mobile phones or vehicles. The user characteristic information refers to the relevant information of the users to be filtered regarding the preset type of commodity. For example, when the preset type of commodity is a mobile phone, the user characteristic information can include the mobile phone brand, model, and advertising messages received by the users to be filtered, etc.

[0047] It should be noted that for the collection of user characteristic information, it can be after obtaining the user's permission to collect the user's characteristic information.

[0048] In some embodiments, the user characteristic information can include at least one of user portrait information, behavior characteristic information, and advertising push characteristic information.

[0049] Among them, the user portrait information can refer to the basic information of the users to be filtered and the relevant information of the users to be filtered currently using regarding the preset type of commodity. For example, the user portrait information can include the province information, city information, age information, gender information, etc. of the users to be filtered. When the preset type of commodity is a mobile phone, the user portrait information can include the mobile phone brand used by the user, the mobile phone signal used, the activation date of the mobile phone used, etc. When the preset type of commodity is a mobile phone, the user portrait information can be as shown in Table 1.

[0050] Table 1

[0051] Serial number Feature Description Dimension 1 Province Province where the user is located 36 2 City City where the user is located 428 3 Mobile phone brand Brand used by the user 16 4 Mobile phone model Model used by the user 1603 5 Activation date Activation date of the mobile phone used by the user 115 6 Inference date Prediction date 115

[0052] It should be understood that the user portrait information shown in Table 1 is used to illustrate the embodiments of the present disclosure and does not limit the specific type of user portrait information. In actual applications, the user portrait information can be set according to actual situations. In addition, for different data dimensions, they are determined according to the corresponding data types.

[0053] Behavior characteristic information refers to the operation statistical information of the user to be screened regarding the advertising messages related to the preset type of products. Among them, the operation statistical information refers to the click behavior of the user to be screened on the advertisements related to the preset type of products. For example, for the advertising messages of mobile phones, the number of clicks on the push (message push) of e-commerce or non-e-commerce platforms can be counted, the number of pushes of the brand of the target product or the brand of non-target products that are clicked, the number of pushes of the brand of the target product contained in the e-commerce platform that are clicked, the number of pushes of the brand of non-target products pushed by the e-commerce platform, the number of pushes of the brand of the target product pushed by the e-commerce platform, the number of pushes of the brand of the target product pushed by the non-e-commerce platform, the number of pushes of the brand of non-target products pushed by the non-e-commerce platform, and so on. When the target product is a Xiaomi mobile phone, the behavior characteristic information can be as shown in Table 2.

[0054] Table 2

[0055]

[0056]

[0057] It should be understood that the behavior characteristic information shown in Table 2 is used to illustrate the embodiments of the present disclosure and does not limit the specific types of behavior characteristic information. In actual applications, the behavior characteristic information can be set according to the actual situation. In addition, for different data dimensions, they are determined according to the corresponding data types.

[0058] Advertising push characteristic information refers to the attribute information of the advertising messages related to the preset type of products received by the user to be screened. Among them, the advertising push characteristic information may include push content (the content of the message push), intent level (the level of intention), feature name (group name), app group (application grouping), app brand (application brand), app type (application type), push brand (the brand of the message push), push brand group (the brand grouping of the message push), push sparse date (message push date), etc. When the target product is a Xiaomi mobile phone, the advertising push characteristic information can be as shown in Table 3.

[0059] Table 3

[0060]

[0061]

[0062] It should be understood that the advertisement push feature information shown in Table 3 is used to illustrate the embodiments of the present disclosure and does not limit the specific types of advertisement push feature information. In actual applications, the advertisement push feature information can be set according to actual situations. Additionally, for different data dimensions, they are determined according to the corresponding data types.

[0063] In step 120, based on the user feature information, determine the target users among the users to be screened who have the target purchase intention, where the target purchase intention refers to the intention of purchasing the target product among the products of the preset type.

[0064] Here, the target users determined from the users to be screened refer to the users with the target purchase intention, where the target purchase intention refers to the intention of purchasing the target product among the products of the preset type. Among them, the intention of purchasing the target product among the products of the preset type can be used for quantitative evaluation to indicate the magnitude of the user's intention to purchase the target product. The greater the intention, the more inclined the user is to purchase the target product. The target product among the products of the preset type refers to a product of a brand within a category of products or a specific product. For example, if the preset type of product is a mobile phone, the target product is the Xiaomi mobile phone or the Xiaomi MIX mobile phone.

[0065] Exemplarily, the user feature information can be used as the input to a trained neural network model to determine the target users from the users to be screened. Or, according to the mapping relationship between the user feature information and the target purchase intention, determine the target users.

[0066] Thus, through multi-dimensional user feature information, the magnitude of the intention of the users to be screened to purchase the target product among the products of the preset type can be accurately predicted, so as to precisely locate potential target users among the users to be screened, provide data support for subsequent marketing strategies, and improve the success rate of marketing strategies. For example, after locating potential target users, methods such as precisely delivering coupons and pushing advertisements can be used to prompt the target users to purchase the target product.

[0067] In some implementable embodiments, the user feature information includes user portrait information, behavioral feature information, and advertisement push feature information. In step 120, the user portrait information, the behavioral feature information, and the advertisement push feature information can be input into a trained product purchase prediction model to determine the target users among the users to be screened who have the target purchase intention.

[0068] Here, the commodity purchase prediction model can be a deep learning model, such as a CNN neural network model. The trained commodity purchase prediction model can be obtained by performing machine learning training on an untrained machine learning model based on user portrait information, behavioral feature information, and advertisement push feature information annotated with scores of target purchase intentions. Input the user portrait information, behavioral feature information, and advertisement push feature information of each user among the users to be screened into the trained commodity purchase prediction model, and the commodity purchase prediction model outputs the score of the user regarding the target purchase intention. This score represents the magnitude of the user's intention to purchase the target commodity among the preset type of commodities, and then determine the target users according to the score. In some examples, users with scores greater than a preset threshold can be determined as target users.

[0069] In some implementable embodiments, the commodity purchase prediction model includes a first sub-model and a second sub-model. Among them, the first sub-model is used to extract a feature vector representing the intention of the user to be screened to purchase the preset type of commodities, and the second sub-model is used to extract a feature vector representing the intention of the user to be screened to purchase the target commodity.

[0070] Figure 2 It is a flowchart for determining target users based on a commodity purchase prediction model shown in an exemplary embodiment, as Figure 2 shown, determining target users based on the trained commodity purchase prediction model may include the following steps.

[0071] In step 221, the user portrait information, the behavioral feature information, and the advertisement push feature information are respectively encoded to obtain a user feature vector, a behavioral feature vector, and an advertisement push feature vector.

[0072] Here, for each feature in the user portrait information, the onehot (one-hot) encoding method can be used for encoding to obtain a user portrait vector. Each feature in the behavioral feature information is a numerical feature, and the normalization method can be used for encoding processing to obtain a behavioral feature vector. For the push content in the advertisement push feature information, the word vector method is used to encode each word in the text, and other features can be encoded using the onehot encoding method to obtain an advertisement push feature vector.

[0073] It should be noted that the dimension of each feature information encoded is the corresponding dimension in Tables 1 to 3 above.

[0074] In step 222, the user portrait information, the behavioral feature information, and the advertisement push feature information are respectively encoded to obtain a user feature vector, a behavioral feature vector, and an advertisement push feature vector.

[0075] Here, the user portrait vector, the behavioral feature vector, and the advertisement push feature vector can be concatenated in sequence to obtain a comprehensive feature vector.

[0076] In step 223, the comprehensive feature vector is used as the input of the first sub-model and the second sub-model respectively, and a first feature vector and a second feature vector are extracted.

[0077] Here, the concatenated comprehensive feature vector is input into the trained first sub-model and second sub-model respectively to obtain a first feature vector and a second feature vector. Among them, the first sub-model and the second sub-model can be CNN convolutional neural network models. The first sub-model is designed to process the comprehensive feature vector into a feature suspension representing the intention of the user to be screened to purchase a preset type of commodity. The second sub-model is designed to process the comprehensive feature vector into an intention representing the user to be screened to purchase the target commodity.

[0078] In step 224, based on the first feature vector and the second feature vector, target users with target purchase intentions among the users to be screened are determined.

[0079] Here, the first feature vector and the second feature vector can be scored respectively, and the scores corresponding to the first feature vector and the second feature vector are used as the magnitudes representing the target purchase intentions of the users to be screened. In some embodiments, the maximum value of the score corresponding to the first feature vector and the score corresponding to the second feature vector can be used as the magnitude of the intention of the user to be screened to purchase the target commodity among the preset types of commodities. When the maximum value of the score corresponding to the first feature vector and the score corresponding to the second feature vector is greater than a preset threshold, the corresponding user to be screened is determined as a target user. For example, a user to be screened with a score greater than 70 is determined as a target user.

[0080] In some implementable embodiments, the method can also concatenate the first feature vector and the second feature vector to obtain a concatenated vector.

[0081] In step 224, determining the intention of the user to purchase the target commodity among the preset types of commodities based on the first feature vector and the second feature vector may include:

[0082] Based on the first feature vector, the second feature vector, and the concatenated vector, target users with target purchase intentions among the users to be screened are determined.

[0083] In some embodiments, the first feature vector can be scored to obtain a first score, the second feature vector can be scored to obtain a second score, and the concatenated vector can be scored to obtain a third score. Then, the user to be screened whose maximum value among the first score, the second score, and the third score is greater than a preset threshold is determined as the target user.

[0084] It should be noted that the commodity purchase prediction model can be obtained by training a machine learning model based on training samples. Among them, the training samples are user feature information marked with scores of target purchase intentions. For example, the target purchase intentions corresponding to the user feature information of different users can be marked to obtain training samples. Then, these training samples are used as the input of an untrained machine learning model, and the machine learning model is trained to obtain a trained commodity purchase prediction model. It should be understood that the commodity purchase prediction model scores the first feature vector, the second feature vector, and the concatenated vector based on the scores corresponding to the feature vectors of different training samples learned by the output layer of the commodity purchase prediction model.

[0085] Among them, the first score represents the magnitude of the intention of the user to be screened to purchase a preset type of commodity. For example, the first score represents the magnitude of the intention of the user to be screened to purchase a mobile phone. The second score represents the magnitude of the intention of the user to be screened to purchase the target commodity. For example, the second score represents the magnitude of the intention of the user to be screened to purchase a Xiaomi mobile phone. The third score represents the magnitude of the intention of the user to be screened to purchase the preset type of commodity and the target commodity among the preset type of commodities. For example, the third score represents the magnitude of the intention of the user to be screened to have both the intention to purchase a mobile phone and the intention to purchase a Xiaomi mobile phone.

[0086] It should be understood that the second score may include users who have not purchased the preset type of commodity but may purchase the target commodity. For example, a user currently has no intention of purchasing a mobile phone, but may choose to purchase a Xiaomi mobile phone when purchasing a mobile phone next time.

[0087] The following combines the attached Figure 3 to elaborate on the above embodiments in detail.

[0088] Figure 3 is an architecture diagram of a commodity purchase prediction model shown according to an exemplary embodiment, as Figure 3As shown in the figure, the commodity purchase prediction model includes an input layer, an encoding layer, a model layer, and an output layer connected in sequence. The model layer includes a first sub-model and a second sub-model. The user portrait information, behavior feature information, and advertisement push feature information are input into the commodity purchase prediction model through the input layer. In the encoding layer, the user portrait information, behavior feature information, and advertisement push feature information are respectively vector-encoded to obtain a user feature vector, a behavior feature vector, and an advertisement push feature vector. And the user feature vector, behavior feature vector, and advertisement push feature vector are concatenated to obtain a comprehensive feature vector. The comprehensive feature vector is respectively used as the input of the first sub-model and the second sub-model of the model layer. The first sub-model and the second sub-model respectively output a first feature vector and a second feature vector, and the first feature vector and the second feature vector are concatenated to obtain a concatenated vector. Then the first feature vector, the concatenated vector, and the second feature vector are respectively used as the input of the output layer, and the output layer outputs a first score, a third score, and a second score. Then, the maximum value among the first score, the third score, and the second score can be used to determine whether the user to be screened is a target user.

[0089] It should be noted that based on the above commodity purchase prediction model, the recall rate can be increased by 9% compared with the traditional single-target and single-task model, and the accuracy rate can be increased by 15%. Among them, the recall rate refers to the ratio of the number of people who purchase mobile phones among the people circled on the same day to the total number of people who purchase mobile phones on the same day, and the accuracy rate refers to the ratio of the number of people who purchase Xiaomi mobile phones to the number of people who purchase mobile phones of all brands among the people circled on the same day.

[0090] Figure 4 It is a block diagram of a device for determining a commodity purchase intention shown according to an exemplary embodiment. Refer to Figure 4 This device includes an acquisition module 401 and an intention determination module 402.

[0091] The acquisition module 401 is configured to determine the user feature information of the user to be screened for the commodity of the preset type;

[0092] The intention determination module 402 is configured to determine the target user with the target purchase intention among the users to be screened based on the user feature information, where the target purchase intention refers to the intention of purchasing the target commodity among the commodities of the preset type.

[0093] In some embodiments, the user feature information includes at least one of the following:

[0094] User portrait information, behavior feature information, and advertisement push feature information, where the behavior feature information represents the operation statistical information of the user for the advertisement messages related to the commodity of the preset type, and the advertisement push feature information represents the attribute information of the advertisement messages related to the commodity of the preset type received by the user.

[0095] In some embodiments, the user feature information includes user profile information, behavioral feature information, and advertisement push feature information; the intention determination module 402 is specifically configured to:

[0096] Input the user profile information, the behavioral feature information, and the advertisement push feature information into a trained product purchase prediction model to determine target users among the users to be screened who have a target purchase intention.

[0097] In some embodiments, the product purchase prediction model includes a first sub-model and a second sub-model. Among them, the first sub-model is used to extract a feature vector representing the intention of the user to be screened to purchase the product of the preset type, and the second sub-model is used to extract a feature vector representing the intention of the user to be screened to purchase the target product;

[0098] The intention determination module 402 includes:

[0099] An encoding unit, configured to encode the user profile information, the behavioral feature information, and the advertisement push feature information respectively to obtain a user feature vector, a behavioral feature vector, and an advertisement push feature vector;

[0100] A first splicing unit, configured to splice the user profile vector, the behavioral feature vector, and the advertisement push feature vector to obtain a comprehensive feature vector;

[0101] A prediction unit, configured to use the comprehensive feature vector as the input of the first sub-model and the second sub-model respectively, and extract a first feature vector and a second feature vector;

[0102] A determination unit, configured to determine target users among the users to be screened who have a target purchase intention based on the first feature vector and the second feature vector.

[0103] In some embodiments, the device further includes:

[0104] A second splicing unit, configured to splice the first feature vector and the second feature vector to obtain a spliced vector;

[0105] The determination unit is specifically configured to:

[0106] Determine target users among the users to be screened who have a target purchase intention based on the first feature vector, the second feature vector, and the spliced vector.

[0107] In some embodiments, the determination unit includes:

[0108] The first scoring unit is configured to determine a first score based on the first feature vector, where the first score represents the magnitude of the intention of the user to be screened to purchase the preset type of commodity;

[0109] The second scoring unit is configured to determine a second score based on the second feature vector, where the second score represents the magnitude of the intention of the user to be screened to purchase the target commodity;

[0110] The third scoring unit is configured to determine a third score based on the concatenated vector, where the third score represents the magnitude of the intention of the user to be screened to purchase the preset type of commodity and the target commodity among the preset type of commodities;

[0111] The purchase intention unit is configured to determine a target user with a target purchase intention among the users to be screened based on the maximum value among the first score, the second score, and the third score.

[0112] In some embodiments, the first sub-model and the second sub-model are CNN convolutional neural network models.

[0113] Regarding the device in the above embodiments, the specific manners in which each module performs operations have been described in detail in the embodiments related to the method, and will not be elaborated herein.

[0114] The present disclosure also provides a computer-readable storage medium, on which computer program instructions are stored, and when the program instructions are executed by a processor, the steps of the method for determining the purchase intention of a commodity provided by the present disclosure are implemented.

[0115] Figure 5 It is a block diagram of an electronic device shown according to an exemplary embodiment. For example, the electronic device 800 may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, etc.

[0116] Refer to Figure 5 , the electronic device 800 may include one or more of the following components: a processing component 802, a memory 804, a power component 806, a multimedia component 808, an audio component 810, an input / output (I / O) interface 812, a sensor component 814, and a communication component 816.

[0117] The processing component 802 generally controls the overall operation of the electronic device 800, such as operations associated with display, telephone calls, data communication, camera operations, and recording operations. The processing component 802 may include one or more processors 820 to execute instructions to complete all or part of the steps of the above-mentioned method for determining the intention of commodity purchase. In addition, the processing component 802 may include one or more modules to facilitate the interaction between the processing component 802 and other components. For example, the processing component 802 may include a multimedia module to facilitate the interaction between the multimedia component 808 and the processing component 802.

[0118] The memory 804 is configured to store various types of data to support the operation of the electronic device 800. Examples of such data include instructions for any application or method operating on the electronic device 800, machine learning models, etc. The memory 804 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disks, or optical disks.

[0119] The power component 806 provides power for various components of the electronic device 800. The power component 806 may include a power management system, one or more power sources, and other components associated with generating, managing, and distributing power for the electronic device 800.

[0120] The multimedia component 808 includes a screen that provides an output interface between the electronic device 800 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen can be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensors can not only sense the boundaries of touch or swipe actions, but also detect the duration and pressure associated with the touch or swipe operation. In some embodiments, the multimedia component 808 includes a front camera and / or a rear camera. When the electronic device 800 is in an operation mode, such as a shooting mode or a video mode, the front camera and / or the rear camera can receive external multimedia data. Each of the front camera and the rear camera can be a fixed optical lens system or have focal length and optical zoom capabilities.

[0121] The audio component 810 is configured to output and / or input audio signals. For example, the audio component 810 includes a microphone (MIC), which is configured to receive external audio signals when the electronic device 800 is in an operating mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signals can be further stored in the memory 804 or transmitted via the communication component 816. In some embodiments, the audio component 810 further includes a speaker for outputting audio signals.

[0122] The I / O interface 812 provides an interface between the processing component 802 and a peripheral interface module, and the peripheral interface module may be a keyboard, a click wheel, buttons, etc. These buttons may include, but are not limited to: a home button, a volume button, a power button, and a lock button.

[0123] The sensor component 814 includes one or more sensors for providing an assessment of various aspects of the status of the electronic device 800. For example, the sensor component 814 can detect the on / off state of the electronic device 800, the relative positioning of components, such as the display and keypad of the electronic device 800. The sensor component 814 can also detect a change in the position of the electronic device 800 or a component of the electronic device 800, the presence or absence of user contact with the electronic device 800, the orientation or acceleration / deceleration of the electronic device 800, and a change in the temperature of the electronic device 800. The sensor component 814 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. The sensor component 814 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor component 814 may further include an acceleration sensor, a gyro sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

[0124] The communication component 816 is configured to facilitate communication between the electronic device 800 and other devices in a wired or wireless manner. The electronic device 800 can access a wireless network based on communication standards, such as WiFi, 2G, or 3G, or a combination thereof. In an exemplary embodiment, the communication component 816 receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 816 further includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.

[0125] In an exemplary embodiment, the electronic device 800 may be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components for performing the above-described method for determining the intention to purchase a commodity.

[0126] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions, such as a memory 804 including instructions, is also provided. The above instructions may be executed by a processor 820 of the electronic device 800 to complete the above-described method for determining the intention to purchase a commodity. For example, the non-transitory computer-readable storage medium may be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, and an optical data storage device, etc.

[0127] In another exemplary embodiment, a computer program product is also provided. The computer program product includes a computer program that can be executed by a programmable device. The computer program has a code portion for performing the above-described method for determining the intention to purchase a commodity when executed by the programmable device.

[0128] Those skilled in the art will readily conceive of other embodiments of the present disclosure after considering the specification and practicing the present disclosure. This application is intended to cover any variations, uses, or adaptations of the present disclosure, which follow the general principles of the present disclosure and include common general knowledge or conventional technical means in the technical field not disclosed in the present disclosure. The specification and embodiments are only to be regarded as exemplary, and the true scope and spirit of the present disclosure are pointed out by the following claims.

[0129] It should be understood that the present disclosure is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present disclosure is only limited by the appended claims.

Claims

1. A method for determining the intention of purchasing goods, characterized in that, it includes: determining the user characteristic information of the users to be screened for goods of a preset type; based on the user characteristic information, determining the target users among the users to be screened who have the target purchase intention, where the target purchase intention refers to the intention of purchasing the target goods among the goods of the preset type; the user characteristic information includes user portrait information, behavior characteristic information, and advertisement push characteristic information; the determining, based on the user characteristic information, the target users among the users to be screened who have the target purchase intention includes: inputting the user portrait information, the behavior characteristic information, and the advertisement push characteristic information into a trained goods purchase prediction model to determine the target users among the users to be screened who have the target purchase intention; the goods purchase prediction model includes an input layer, an encoding layer, a model layer, and an output layer connected in sequence. Through the input layer, the user portrait information, the behavior characteristic information, and the advertisement push characteristic information are input into the goods purchase prediction model. In the encoding layer, vector encoding is respectively performed on the user portrait information, the behavior characteristic information, and the advertisement push characteristic information to obtain a user characteristic vector, a behavior characteristic vector, and an advertisement push characteristic vector. The user characteristic vector, the behavior characteristic vector, and the advertisement push characteristic vector are concatenated to obtain a comprehensive characteristic vector. The comprehensive characteristic vector is respectively used as the input of the first sub-model and the second sub-model in the model layer to obtain the target users among the users to be screened who have the target purchase intention; the first sub-model and the second sub-model are used to extract a first characteristic vector and a second characteristic vector from the comprehensive characteristic vector; the output layer is used to determine a first score based on the first characteristic vector, where the first score represents the size of the intention of the user to be screened to purchase the goods of the preset type; determine a second score based on the second characteristic vector, where the second score represents the size of the intention of the user to be screened to purchase the target goods; determine a third score based on the concatenated vector, where the third score represents the size of the intention of the user to be screened to purchase the goods of the preset type and purchase the target goods among the goods of the preset type at the same time. The concatenated vector is obtained by concatenating the first characteristic vector and the second characteristic vector; based on the maximum value among the first score, the second score, and the third score, determine the target users among the users to be screened who have the target purchase intention.

2. The method for determining the intention of purchasing goods according to claim 1, characterized in that, the behavior characteristic information represents the operation statistical information of the user for the advertisement messages related to the goods of the preset type, and the advertisement push characteristic information represents the attribute information of the advertisement messages related to the goods of the preset type received by the user.

3. The method for determining the intention of purchasing goods according to claim 2, characterized in that, The first sub-model is used to extract a feature vector characterizing the intention of the user to be screened to purchase the preset type of commodity, and the second sub-model is used to extract a feature vector characterizing the intention of the user to be screened to purchase the target commodity.

4. The method for determining the commodity purchase intention according to claim 3, wherein, the first sub-model and the second sub-model are CNN convolutional neural network models.

5. A device for determining the commodity purchase intention, wherein, it includes: an acquisition module configured to determine user feature information of a user to be screened for a preset type of commodity; an intention determination module configured to determine a target user with a target purchase intention among the users to be screened based on the user feature information, where the target purchase intention refers to the intention to purchase the target commodity among the preset type of commodities; the user feature information includes user portrait information, behavioral feature information, and advertisement push feature information; the intention determination module is specifically configured to: input the user portrait information, the behavioral feature information, and the advertisement push feature information into a trained commodity purchase prediction model to determine a target user with a target purchase intention among the users to be screened; the commodity purchase prediction model includes an input layer, an encoding layer, a model layer, and an output layer connected in sequence. The user portrait information, the behavioral feature information, and the advertisement push feature information are input into the commodity purchase prediction model through the input layer. In the encoding layer, vector encoding is respectively performed on the user portrait information, the behavioral feature information, and the advertisement push feature information to obtain a user feature vector, a behavioral feature vector, and an advertisement push feature vector. The user feature vector, the behavioral feature vector, and the advertisement push feature vector are concatenated to obtain a comprehensive feature vector. The comprehensive feature vector is respectively used as the input of the first sub-model and the second sub-model of the model layer to obtain a target user with a target purchase intention among the users to be screened; the first sub-model and the second sub-model are used to extract a first feature vector and a second feature vector from the comprehensive feature vector; the output layer is configured to determine a first score based on the first feature vector, where the first score characterizes the magnitude of the intention of the user to be screened to purchase the preset type of commodity; determine a second score based on the second feature vector, where the second score characterizes the magnitude of the intention of the user to be screened to purchase the target commodity; determine a third score based on a concatenated vector, where the third score characterizes the magnitude of the intention of the user to be screened to purchase the preset type of commodity and the target commodity among the preset type of commodities at the same time. The concatenated vector is obtained by concatenating the first feature vector and the second feature vector; determine a target user with a target purchase intention among the users to be screened based on the maximum value of the first score, the second score, and the third score.

6. A computer-readable storage medium, on which computer program instructions are stored, wherein, When the program instruction is executed by a processor, it implements the steps of the method according to any one of claims 1 to 4.

7. An electronic device, characterized in that it comprises: a memory storing a computer program thereon; a processor configured to execute the computer program in the memory to implement the steps of the method according to any one of claims 1 to 4.

Images