CN114912009A - User portrait generation method, device, electronic equipment and computer program medium - Google Patents

Abstract

The embodiment of the application provides a user portrait generation method and device, electronic equipment and a computer program medium, and relates to the technical field of artificial intelligence. The method in the embodiment of the application comprises the following steps: acquiring user characteristic information of a target user; respectively generating a plurality of user characteristic vectors of a target user under a user characteristic dimension based on the user characteristic information; inputting a plurality of user feature vectors into a pre-trained machine learning model; acquiring preference level labels of target objects under the target classification attributes of target users output by a pre-trained machine learning model; and if the acquired preference level label is matched with a preset preference level label, generating a user portrait of the target user based on the target classification attribute. The technical scheme of the embodiment of the application improves the accuracy of the obtained user portrait.

Description

User portrait generation method, apparatus, electronic device and computer program medium

technical field

The present application relates to the field of computer technology, and in particular, to a method, apparatus, electronic device, and computer program medium for generating a user portrait.

Background technique

With the development of Internet technology, obtaining user portraits based on big data, and then realizing various businesses through user portraits has become one of the core technologies of the Internet. Attributes, interests and preferences and other dimensions of data, accurately describe and analyze all aspects of user information, and tap potential value.

In the related art, when a user portrait of a user is generated, the portrait label is generally extracted from the user behavior data, and the portrait labels involved in the user behavior data are counted, and each user's portrait label is scored according to the frequency of statistics. Then, the user portrait is obtained according to the score of the portrait label. For users who have not generated behaviors, due to the lack of behavioral data, it is difficult to obtain user portraits based on tag statistics or the accuracy of user portraits obtained based on tag statistics is low, which in turn affects the performance of related services based on user portraits. precision.

SUMMARY OF THE INVENTION

The embodiments of the present application provide a method, apparatus, electronic device, and computer program medium for generating a user portrait, which are used to improve the accuracy of the obtained user portrait.

Other features and advantages of the present application will become apparent from the following detailed description, or be learned in part by practice of the present application.

According to an aspect of the embodiments of the present application, a method for generating a user portrait is provided, including: acquiring user feature information of a target user, where the user feature information includes user attribute information and historical behavior data of the target user; the user feature information, and respectively generate multiple user feature vectors of the target user under the user feature dimension, where the user feature dimension refers to the category of the user attribute information and the operation events contained in the historical behavior data The feature dimension composed of the category and the classification attribute of the operation object; the multiple user feature vectors are input into the pre-trained machine learning model, and the pre-trained machine learning model contains the sample users under the user feature dimension. a user feature vector and the preference grade label of the sample user for the target object under the target classification attribute; obtain the target user's preference grade label for the target object under the target classification attribute output by the pre-trained machine learning model; if the obtained preference If the grade label matches the preset preference grade label, the user portrait of the target user is generated based on the target classification attribute.

According to an aspect of the embodiments of the present application, an apparatus for generating a user portrait is provided, including: a first obtaining unit, configured to obtain user feature information of a target user, where the user feature information includes user attribute information of the target user and historical behavior data; the first generating unit is used to respectively generate a plurality of user feature vectors of the target user under the user feature dimension based on the user feature information, and the user feature dimension refers to the user attribute information determined by the user feature information. category, and the feature dimension composed of the operation event category included in the historical behavior data and the classification attribute of the operation object; the input unit is used to input the multiple user feature vectors into the pre-trained machine learning model, the The pre-trained machine learning model contains multiple user feature vectors of the sample user under the user feature dimension and the sample user's preference level label for the target object under the target classification attribute; the output unit is used to obtain the pre-trained data. The preference grade label of the target user for the target object under the target classification attribute output by the machine learning model; the second generating unit is used for if the obtained preference grade label matches the preset preference grade label, based on the target classification attribute A user portrait of the target user is generated.

In some embodiments of the present application, based on the foregoing solution, the device for generating user portraits further includes: a third generating unit, configured to perform conversion processing on the plurality of user feature vectors respectively to generate the plurality of user features The first feature vector corresponding to the vector, the first feature vector corresponding to the multiple user features is a vector of the same dimension; the aggregation unit is used for the preset association relationship between the multiple user feature information. The first feature vectors corresponding to the plurality of user feature vectors are aggregated to generate a plurality of second feature vectors after aggregation; a prediction unit is used to predict the target user's response to the target based on the plurality of second feature vectors after the aggregation. The preference level label of the target object under the classification attribute, and the first feature vector corresponding to the multiple user features is a vector of the same dimension.

In some embodiments of the present application, based on the foregoing solution, the device for generating user portraits further includes: a second acquiring unit, configured to acquire training set sample data for training the machine learning model to be trained, and the training Each piece of sample data in the set sample data includes multiple user feature vectors of the sample user under the user feature dimension and the sample user's preference level label for the target object under the target classification attribute; the training unit is used to pass the training set sample The data is used to train the machine learning model to be trained to obtain a pre-trained machine learning model.

In some embodiments of the present application, based on the foregoing solution, the second obtaining unit is configured to: obtain user feature information of candidate users in the candidate user set; for each candidate user in the candidate user set, based on the candidate user The historical behavior data in the user feature information of the A candidate user of an operation event, the target operation event refers to an operation event that operates the target object under the target classification attribute; based on the user feature information of the first type of candidate user, the first type of candidate user is generated Multiple user feature vectors under the user feature dimension, and add preference level labels with high preference to the first type of candidate users to obtain sample data of positive sample users; based on the user characteristic information of the second type of candidate users , generate a plurality of user feature vectors of the second type of candidate users under the user characteristic dimension, and add a preference level label with a low degree of preference to the second type of candidate users to obtain sample data of negative sample users; based on the The sample data of the positive sample user and the sample data of the negative sample user are used to obtain the training set sample data for training the machine learning model to be trained.

In some embodiments of the present application, based on the foregoing solution, the second obtaining unit is configured to: extract candidate users from the second type of candidate users; generate the extracted candidate users based on user feature information of the extracted candidate users Multiple user feature vectors of the candidate users under the user feature dimension, and adding a preference level label with a low degree of preference to the extracted candidate users to obtain sample data of negative sample users.

In some embodiments of the present application, based on the foregoing solution, the second acquisition unit is configured to: determine the second type of candidate user based on the frequency of historical behavior data included in the user feature information of the second type of candidate user The portrait density is positively correlated with the frequency; in the second type of candidate users whose portrait density is higher than the predetermined portrait density threshold, the candidate users are extracted at the first ratio, and the Among the candidate users of the second category whose portrait density is lower than or equal to a predetermined portrait density threshold, candidate users are extracted with a second proportion, and the first proportion is greater than the second proportion.

In some embodiments of the present application, based on the foregoing solution, the second obtaining unit is configured to: obtain a configuration file, where the configuration file is used to identify the attribute category of the obtained user attribute information and the obtained historical behavior data. The operation event category and the classification attribute of the operation object are configured; for each candidate user in the candidate user set, based on the category of the user attribute information configured in the configuration file, the user of the candidate user under the configured attribute category is obtained. Attribute information; for each candidate user in the candidate user set, based on the operation event category and the classification attribute of the operation object configured in the configuration file, obtain the candidate user's information under the operation event category and the classification attribute of the operation object. Historical behavioral data.

According to an aspect of the embodiments of the present application, a computer-readable medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, implements the method for generating a user portrait as described in the foregoing embodiments.

According to an aspect of the embodiments of the present application, an electronic device is provided, including: one or more processors; and a storage device for storing one or more programs, when the one or more programs are stored by the one or more programs When executed by multiple processors, the one or more processors are made to implement the method for generating a user portrait as described in the foregoing embodiments.

According to one aspect of the embodiments of the present application, there is provided a computer program product or computer program, where the computer program product or computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the method for generating a user portrait provided in the above-mentioned various optional embodiments.

In the technical solutions provided by some embodiments of the present application, user feature information of the target user is obtained, and the user feature information includes the user attribute information and historical behavior data of the target user; based on the user feature information, the user feature dimension of the target user is generated respectively Multiple user feature vectors below, the user feature dimension refers to the feature dimension composed of the category of user attribute information, the operation event category included in the historical behavior data, and the classification attribute of the operation object; In the trained machine learning model, the pre-trained machine learning model obtains the pre-trained machine by including multiple user feature vectors of the sample user under the user feature dimension and the sample user's preference level label for the target object under the target classification attribute; The preference grade label of the target user for the target object under the target classification attribute output by the learning model; if the obtained preference grade label matches the preset preference grade label, a user portrait of the target user is generated based on the target classification attribute. Compared with obtaining user portraits based on tag statistics, user feature vectors can more comprehensively represent user preferences, which can improve the accuracy of predicting users' preference grade labels for target objects under specific classification attributes, thereby facilitating the identification of specific users. The target users of the portrait, in order to improve the accuracy of the relevant business recommendation.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not limiting of the present application.

Description of drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the application and together with the description serve to explain the principles of the application. Obviously, the drawings in the following description are only some embodiments of the present application, and for those of ordinary skill in the art, other drawings can also be obtained from these drawings without creative effort. In the attached image:

FIG. 1 shows a schematic diagram of an exemplary system architecture to which the technical solutions of the embodiments of the present application can be applied.

FIG. 2 shows a flowchart of a method for generating a user portrait according to an embodiment of the present application.

FIG. 3 shows a flowchart of a method for generating a user portrait according to an embodiment of the present application.

FIG. 4 shows a schematic structural diagram of a pre-trained machine learning model according to an embodiment of the present application.

FIG. 5 shows a schematic diagram of a specific structure of a capsule layer in a pre-trained machine learning model according to an embodiment of the present application.

FIG. 6 shows a flowchart of a method for generating a user portrait according to an embodiment of the present application.

FIG. 7 shows a specific flowchart of step S610 of the method for generating a user portrait according to an embodiment of the present application.

FIG. 8 shows a specific flowchart of step S710 of the method for generating a user portrait according to an embodiment of the present application.

FIG. 9 shows a specific flowchart of step S740 of the method for generating a user portrait according to an embodiment of the present application.

FIG. 10 shows a specific flowchart of step S910 of the method for generating a user portrait according to an embodiment of the present application.

FIG. 11 shows a block diagram of an apparatus for generating a user portrait according to an embodiment of the present application.

FIG. 12 shows a schematic structural diagram of a computer system suitable for implementing the electronic device according to the embodiment of the present application.

Detailed ways

Example embodiments will now be described more fully with reference to the accompanying drawings. Example embodiments, however, can be embodied in various forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this application will be thorough and complete, and will fully convey the concept of example embodiments to those skilled in the art.

Furthermore, the described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided in order to give a thorough understanding of the embodiments of the present application. However, those skilled in the art will appreciate that the technical solutions of the present application may be practiced without one or more of the specific details, or other methods, components, devices, steps, etc. may be employed. In other instances, well-known methods, devices, implementations, or operations have not been shown or described in detail to avoid obscuring aspects of the present application.

The block diagrams shown in the figures are merely functional entities and do not necessarily necessarily correspond to physically separate entities. That is, these functional entities may be implemented in software, or in one or more hardware modules or integrated circuits, or in different networks and/or processor devices and/or microcontroller devices entity.

The flowcharts shown in the figures are only exemplary illustrations and do not necessarily include all contents and operations/steps, nor do they have to be performed in the order described. For example, some operations/steps can be decomposed, and some operations/steps can be combined or partially combined, so the actual execution order may be changed according to the actual situation.

Artificial Intelligence (AI): It is a theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technique of computer science that attempts to understand the essence of intelligence and produce a new kind of intelligent machine that can respond in a similar way to human intelligence. Artificial intelligence is to study the design principles and implementation methods of various intelligent machines, so that the machines have the functions of perception, reasoning and decision-making.

Artificial intelligence technology is a comprehensive discipline, involving a wide range of fields, including both hardware-level technology and software-level technology. The basic technologies of artificial intelligence generally include technologies such as sensors, special artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation/interaction systems, and mechatronics. Artificial intelligence software technology mainly includes computer vision technology, speech processing technology, natural language processing technology, and machine learning/deep learning. For example, in the embodiment of the present application, the artificial intelligence technology is used to determine multiple user feature vectors of the target user under the user feature dimension, and based on the multiple user feature vectors of the target user under the user feature dimension, the target user is determined to classify the target The preference grade label of the target object under the attribute, and then the preference grade label of the target user to the target object under the target classification attribute, and the user portrait of the target user is generated.

User portrait: A user portrait is a labelled user model abstracted from information such as user social attributes, living habits and consumption behavior. The core work of building user portraits is to “label” users, and labels are highly refined feature identifiers obtained by analyzing user information.

Machine Learning (ML, Machine Learning) is a multi-field interdisciplinary subject involving probability theory, statistics, approximation theory, convex analysis, algorithm complexity theory and other disciplines. It specializes in how computers simulate or realize human learning behaviors to acquire new knowledge or skills, and to reorganize existing knowledge structures to continuously improve their performance. Machine learning is the core of artificial intelligence and the fundamental way to make computers intelligent, and its applications are in all fields of artificial intelligence. Machine learning and deep learning usually include artificial neural networks, belief networks, reinforcement learning, transfer learning, inductive learning, teaching learning and other technologies.

FIG. 1 shows a schematic diagram of an exemplary system architecture to which the technical solutions of the embodiments of the present application can be applied.

As shown in FIG. 1 , the system architecture may include a user profile demander platform 101 , a network 102 and a user profile provider platform 103 . The user profile demander platform 101 and the user profile provider platform 103 are connected through a network 102, and data interaction is performed based on the network 102. The network may include various connection types, such as wired communication links, wireless communication links and so on.

It should be understood that the numbers of user portrait demander platforms 101 , networks 102 and user portrait provider platforms 103 in FIG. 1 are merely illustrative. According to implementation requirements, there may be any number of user portrait demander platforms 101 , networks 102 and user portraiture provider platforms 103 . For example, the user portrait provider platform 103 can be a server cluster that provides user portrait generation services, the user portrait demander platform 101 can be a server cluster or client that needs to obtain user portraits, and the client can be mobile phones, tablets, portable computers and desktops One or more kinds of computers, of course, are not limited to this. The user portrait provider platform 103 may be a platform that provides users with various business services, such as social applications or instant messaging applications, which may include user attribute information and historical behavior data of a large number of users.

The user portrait demander platform 101 provides the identification information of the target user of the user portrait that needs to be generated, and the user portrait provider platform 103 obtains the user characteristic information of the target user based on the user identification information, and the user characteristic information includes the user attribute information and history of the target user. Behavior data; based on the user feature information, generate multiple user feature vectors of the target user under the user feature dimension. The user feature dimension refers to the category of the user feature information, as well as the operation event category and operation object contained in the historical behavior data. The feature dimension composed of categorical attributes; multiple user feature vectors are input into the pre-trained machine learning model, and the pre-trained machine learning model contains multiple user feature vectors of the sample users under the user feature dimension and the sample user's target The preference level label of the target object under the classification attribute; obtain the preference level label of the target user for the target object under the target classification attribute output by the pre-trained machine learning model; if the obtained preference level label matches the preset preference level label , the user portrait of the target user is generated based on the target classification attribute.

It can be seen from the above that, compared with obtaining user portraits based on tag statistics, user feature vectors can more comprehensively represent user preferences, which can improve the accuracy of predicting users’ preference grade labels for target objects under specific classification attributes, which is convenient for Find out target users with specific user profiles, so as to improve the accuracy of relevant business recommendations.

It should be noted that the method for generating the user portrait provided in the embodiment of the present application is generally performed by the user portrait provider platform 103 , and correspondingly, the user portrait generating device is generally set in the user portrait provider platform 103 . However, in other embodiments of the present application, the user portrait demander platform 101 may also have similar functions to the user portrait provider platform 103, so as to implement the solution of the user portrait generation method provided by the embodiments of the present application. The implementation details of the technical solutions of the embodiments of the present application are described in detail below.

FIG. 2 shows a flowchart of a method for generating a user portrait according to an embodiment of the present application. The method for generating a user portrait may be performed by a user portrait provider platform, and the server may be the user portrait shown in FIG. 1 . Provider Platform 103. Referring to FIG. 2 , the method for generating a user portrait includes at least steps S210 to S250, which are described in detail as follows.

In step S210, user characteristic information of the target user is acquired, where the user characteristic information includes user attribute information and historical behavior data of the target user.

In an embodiment of the present application, the user feature information refers to the feature information of the user in multiple dimensions, which specifically includes user attribute information and historical behavior data, and the target user refers to the user who needs to generate a user portrait.

In an embodiment of the present application, the user characteristic information of the target user can be obtained from the registration information of the target user by the user portrait provider platform, and the user attribute information of the target user can include various types of attribute information, such as age, birth Date, gender, address and position, etc., of course, are not limited to this.

In an embodiment of the present application, the historical behavior data of the target user includes the historical behavior data of the target user in the user portrait provider platform. The behavior data includes operation events and attribute information of the operation objects, and the operation events can be click, browse, favorite, comment, forward, etc. The operation object may be commodity advertisement or content, etc., and the classification attribute of the operation object may include subject, category, label, and so on.

Optionally, when the operation object is a commodity advertisement, the operation event can be a browsing event, and its classification attribute can be a category; when the operation object is content, the operation event can be a click event, a browsing event, a favorite event, a comment event, a forwarding event, etc. Events, whose classification attributes can be topics, categories, and tags.

In step S220, based on the user feature information, respectively generate multiple user feature vectors of the target user under the user feature dimension, where the user feature dimension refers to the category of the user attribute information and the operation event category and operation included in the historical behavior data The feature dimension formed by the categorical attributes of an object.

In an embodiment of the present application, the user feature dimension refers to a feature dimension composed of categories of user attribute information, operation event categories included in historical behavior data, and classification attributes of operation objects.

In an embodiment of the present application, for the feature dimension composed of categories of user attribute information, each category of user attribute information will generate a user feature vector. If the user attribute information includes age and gender, then age and gender constitute Two feature dimensions, thus generating a user feature vector based on age and gender.

In an embodiment of the present application, for the feature dimension formed by the operation event category included in the historical behavior data and the classification attribute of the operation object, it can be implemented in various ways.

In one embodiment, the feature dimension may be determined according to the classification attribute of the operation object. Taking the operation object as content as an example, when the classification attributes of the content include subject, category, and tag, then the subject, category, and tag are taken as one feature dimension respectively, and then user feature vectors of three feature dimensions are generated.

In another embodiment, the feature dimension may be jointly determined according to the classification attribute of the operation object and the operation event category. Taking the operation object as content as an example, when the classification attributes of the content include topic, category, and label, and the operation event category includes click events and comment events, then the click event for the topic, the comment event for the topic, and the category The click event, the comment event for the category, the click event for the label, and the comment event for the label are respectively taken as a feature dimension, and then a user feature vector of six feature dimensions is generated.

It should be noted that the implementation of the feature dimension division is not limited to the above-mentioned implementation, and other implementations may also be used, which are not specifically limited in this application.

In step S230, a plurality of user feature vectors are input into a pre-trained machine learning model, and the pre-trained machine learning model includes a plurality of user feature vectors of the sample user under the user feature dimension and the sample user's target classification attribute Target object's preference rating under the tab.

In an embodiment of the present application, multiple user feature vectors of the target user under the user feature dimension can be input into a pre-trained machine learning model, and the pre-trained machine learning model will assign the target user to the target under the target classification attribute. The preference level label of the object, where the pre-trained machine learning model may be a look-alike model, etc., or a deep neural network model or a CNN (Convolutional Neural Network, convolutional neural network) model. The preference level label can reflect the preference degree of the target user to the target object under the target classification attribute. The preference level label can include two levels of "interested" and "not interested" from high to low. Specifically, the preference level label is from high to low. Low can also include four levels of "special interest", "interested", "not interested" and "especially not interested". The higher the level, the higher the target user's preference for the target object under the target classification attribute.

Referring to FIG. 3, FIG. 3 shows a flow chart of a method for generating a user portrait according to an embodiment of the present application. The pre-trained machine learning model determines the preference level of the target user to the target object under the target classification attribute based on the following method The label may specifically include steps S310 to S330.

Referring to FIG. 4, FIG. 4 shows a schematic structural diagram of a pre-trained machine learning model according to an embodiment of the present application, and the pre-trained machine learning model shown in FIG. 4 may specifically include a projection layer (Projection Layer) 402, Capsule Layer 403 and Capsule Attention Layer 405.

Referring to FIG. 5 , FIG. 5 shows a specific structural schematic diagram of a capsule layer in a pre-trained machine learning model according to an embodiment of the present application. Steps S310 to S330 are described in detail below with reference to 3 to FIG. 5 .

In step S310, the conversion processing is performed on the plurality of user feature vectors respectively to generate first feature vectors corresponding to the plurality of user feature vectors, and the first feature vectors corresponding to the plurality of user features are vectors of the same dimension.

In an embodiment of the present application, as shown in FIG. 4 , after multiple user feature vectors 401 are input into the pre-trained machine learning model, it converts the multiple user feature vectors through a projection layer (Projection Layer) 402 processing to generate first feature vectors corresponding to multiple user feature vectors. It should be noted that the first feature vectors corresponding to multiple user features are vectors of the same dimension.

为第i个用户特征向量，

为第i个用户特征向量对应的第一特征向量，e和m为固定参数，W_d为对对多个用户特征向量进行转换处理的转换矩阵。通过投影层(Projection Layer)将多个用户特征向量转换成同一维度的向量，可以提高预训练的机器学习模型的数据处理效率。Specifically, when the projection layer (Projection Layer) 402 is used to convert multiple user feature vectors to generate the first feature vector corresponding to the multiple user feature vectors, the formula can be used to achieve

is the i-th user feature vector,

is the first eigenvector corresponding to the i-th user eigenvector, e and m are fixed parameters, and W _d is a conversion matrix for converting multiple user eigenvectors. Converting multiple user feature vectors into vectors of the same dimension through the projection layer can improve the data processing efficiency of the pre-trained machine learning model.

In step S320, based on the preset association relationship between multiple user feature information, an aggregation process is performed on the first feature vectors corresponding to the multiple user feature vectors to generate multiple aggregated second feature vectors, and multiple user feature vectors are generated. The first feature vector corresponding to the feature is a vector of the same dimension.

In an embodiment of the present application, for the first feature vector corresponding to the plurality of user feature vectors obtained through the projection layer (Projection Layer), the capsule layer (Capsule Layer) 403 is based on the association relationship between the plurality of user feature information, Aggregate the first feature vectors corresponding to multiple user feature vectors to generate multiple second feature vectors after aggregation, and the first feature vectors corresponding to multiple user features are vectors of the same dimension

可以是根据预定的公式

对其进行聚合处理，生成对应的向量

并对

通过公式

进行归一化处理生成单位向量

Specifically, for the i-th first feature vector input into the Capsule Layer 403

can be based on a predetermined formula

Aggregate it to generate the corresponding vector

and to

by formula

Perform normalization to generate unit vectors

来更新的，

为是对

的转置变换得到的，当待训练的机器学习模型达到收敛条件时，则第一特征向量进行聚合时的权重w_ij也随即确定，Squash是对

进行归一化处理的转换矩阵。I _u represents the sequence number of the first eigenvector, S is the transformation matrix, w _ij is the weight when the i-th first eigenvector and the j-th first eigenvector are aggregated, and b _ij is the capsule layer (Capsule Layer ) 403 is the parameter initialized when performing forward propagation calculation, and w _ij may be determined by w _ij ←softmax(b _ij ), that is, by normalizing b _ij to obtain w _ij . It should be pointed out that when the machine learning model to be trained is trained, when the parameters in the machine learning model to be trained are updated by forward propagation calculation, b _ij is based on the formula

to update,

for yes

obtained from the _transpose transformation of

The transformation matrix to be normalized.

In step S330, based on the aggregated multiple second feature vectors, the preference level label of the target user for the target object under the target classification attribute is predicted.

In one embodiment of the present application, the Capsule Attention Layer 405 analyzes and processes the aggregated multiple second feature vectors, predicts the target user's preference score for the target object under the target classification attribute, and predicts the target user's preference score according to the target classification attribute. The correspondence between the preference score score and the preference grade label predicts the target user's preference grade label for the target object under the target classification attribute. It can be understood that the higher the preference score, the higher the corresponding preference grade label.

计算来得到高阶特征交叉

并通过公式

来对

进行归一化处理，得到归一化处理后的a_ij，基于a_ij来预测目标用户对目标分类属性下的目标对象的偏好得分score。其中，b为固定参数，W∈R^t×k，b∈R^t，h∈R^t，t表示胶囊注意层405的隐层维度，k表示的是第二特征向量的向量维度，h^T表示的是输入的聚合后的多个第二特征向量的集合，v_i和v_j分别为聚合后的多个第二特征向量的集合中的第i个第二特征向量与第j个第二特征向量，v_i·v_j为第i个第二特征向量与第j个第二特征向量之间的数量积。Specifically, when the capsule attention layer 405 performs aggregation processing on the input aggregated second feature vectors, the specific method can firstly be performed according to the formula

Compute to get higher-order feature intersections

and by formula

come right

Perform normalization processing to obtain a _ij after normalization processing, and predict the target user's preference score score for the target object under the target classification attribute based on a _ij . Among them, b is a fixed parameter, W∈R ^t×k , b∈R ^t , h∈R ^t , t denotes the hidden layer dimension of the capsule attention layer 405 , k denotes the vector dimension of the second feature vector, h ^T denotes is the input set of multiple second feature vectors after aggregation, v _i and v _j are the i-th second feature vector and the j-th second feature in the set of multiple second feature vectors after aggregation, respectively vector, v _i ·v _j is the quantity product between the ith second eigenvector and the jth second eigenvector.

Referring to FIG. 6, FIG. 6 shows a flowchart of a method for generating a user portrait according to an embodiment of the present application. The method for generating a user portrait in this embodiment may further include steps S610 to S620, which are described in detail as follows.

In step S610, a training set sample data for training the machine learning model to be trained is obtained, and each piece of sample data in the training set sample data includes multiple user feature vectors of the sample user under the user feature dimension and a pair of sample user pairs. The preference rating label of the target object under the target classification attribute.

In an embodiment of the present application, the training set sample data includes a large number of sample data for training a machine learning model to be trained. When generating the sample data in the sample data of the training set, the existing users in the user portrait provider platform can be used as sample users, and the user attribute information and historical behavior data of the sample users can be obtained by obtaining the user attribute information and The historical behavior data respectively generates multiple user feature vectors of the sample users under the user feature dimension. In addition, for each sample user, it is also necessary to determine the target classification attribute of the sample user according to its specific user attribute information and historical behavior data. The preference status of the target object, and then the preference level label of the sample user for the target object under the target classification attribute is generated according to the determined preference status.

Referring to FIG. 7 , FIG. 7 shows a specific flowchart of step S610 of a method for generating a user portrait according to an embodiment of the present application. The step S610 may include steps S710 to S750 , which are described in detail as follows.

In step S710, user feature information of the candidate users in the candidate user set is obtained.

In one embodiment, the candidate user set is an existing user set in the user profile provider platform, which includes a large number of candidate users that can be selected as sample users.

Referring to FIG. 8 , FIG. 8 shows a specific flowchart of step S710 of a method for generating a user portrait according to an embodiment of the present application. The step S710 may include steps S810 to S830 , which are described in detail as follows.

In step S810, a configuration file is obtained, where the configuration file is used to configure the attribute category of the acquired user attribute information, the operation event category of the acquired historical behavior data, and the classification attribute of the operation object.

In an embodiment of the present application, in order to obtain user feature information of candidate users, in order to facilitate the generation of a specific user portrait according to specific business scenario requirements, the attribute category of the user attribute information to be obtained and the historical behavior data can be pre-determined. Configure the operation event category and the classification attribute of the operation object to generate a corresponding configuration file.

Specifically, when the user portrait to be generated is for a product advertisement, in the configuration file, the attribute category of the user attribute information to be obtained can include age, gender and address, and the operation event category of the historical behavior data can be click For events and browsing events, the classification attribute of the operation object may include the product category in the product advertisement. It should be noted that the product category may include one or more levels of classification categories. Taking the commodity category including the first-level commodity category and the second-level commodity category as an example, the first-level commodity category can include home appliances, food and beverages, medicine and health care, etc., while the second-level commodity category is the specific category within the first-level commodity category, such as home appliances The secondary commodity categories of a tv can include TVs, air conditioners, refrigerators, etc.

In step S820, for each candidate user in the candidate user set, based on the category of the user attribute information configured in the configuration file, user attribute information of the candidate user under the configured attribute category is obtained.

In an embodiment of the present application, after the configuration file is obtained, for each candidate user in the candidate user set, based on the configured user attribute information category in the configuration file, the configured attribute category of the candidate user in the configured attribute category can be obtained respectively. User attribute information under .

In step S830, for each candidate user in the candidate user set, based on the operation event category and the classification attribute of the operation object configured in the configuration file, the historical behavior data of the candidate user under the operation event category and the classification attribute of the operation object is obtained.

In an embodiment of the present application, after obtaining the configuration file, for each candidate user in the candidate user set, the corresponding operation event category and the classification attribute of the operation object configured in the configuration file can be obtained. Historical behavior data under the action event category and the action object's classification attribute.

In the technical solution of the embodiment shown in FIG. 8, through the configuration file, the attribute category of the user attribute information to be acquired, the operation event category of the historical behavior data to be acquired, and the classification attribute of the operation object can be configured, and further Obtaining user characteristic data of a specific dimension according to the requirements can obtain the required characteristic data in a targeted manner, thereby facilitating the generation of sample data that conforms to a specific business scenario.

Please also continue to refer to FIG. 7, in step S720, for each candidate user in the candidate user set, based on the historical behavior data in the user feature information of the candidate user, determine the first type of candidate users and the second type of candidate users, the first type of candidate user. The class candidate user is a candidate user that includes the target operation event, the second class candidate user is a candidate user that does not include the target operation event, and the target operation event refers to the operation event that operates the target object under the target classification attribute.

In an embodiment of the present application, since the pre-trained machine learning model is used to predict the user's preference level label for the target object under the specific target classification attribute, the sample users for training the machine learning model may include positive sample users and negative sample users, positive sample users are sample users with higher preference for the target object under the specific target classification attribute, and negative sample users have lower preference for the target object under the specific target classification attribute. The sample data composed of sample users and negative sample users is used to train the machine learning model, so that the machine learning model can effectively identify whether the target user is a user with a high degree of preference for the target object under the specific target classification attribute.

In an embodiment of the present application, the target operation event refers to an operation event that the user operates on the target object under the target category attribute, such as a click event, browsing event, etc. for the target object under the target category attribute. When the user's historical behavior data includes the target operation event, it can indicate that the degree of interest in the target object under the target classification attribute is high; otherwise, it indicates that the interest degree in the target object under the target classification attribute is low. Therefore, for each candidate user in the candidate user set, the candidate users can be classified according to whether the historical behavior data of the candidate user includes the target operation event, and the first type of candidate users including the target operation event candidate user and the candidate users not including the target operation event can be obtained. For the second type of candidate users of the candidate users of the operation event, it can be understood that the first type of candidate users is used as a candidate user for generating positive sample users, and the second type of candidate users is used as a candidate user for generating negative sample users.

In step S730, based on the user characteristic information of the first type of candidate users, multiple user characteristic vectors of the first type of candidate users under the user characteristic dimension are generated, and a preference level label with a high degree of preference is added to the first type of candidate users, Get sample data of positive sample users.

In an embodiment of the present application, when generating the sample data of the positive sample users, based on the user feature information of the first-type candidate users, multiple user feature vectors of the first-type candidate users in the user feature dimension can be generated, and Add preference grade labels with high preference for the first type of candidate users.

Optionally, for the first type of candidate users, all candidate users in the first type of candidate users may be selected as positive sample users.

Optionally, for the first type of candidate users, some candidate users in the first type of candidate users may be selected as positive sample users.

In step S740, based on the user characteristic information of the second type of candidate users, multiple user characteristic vectors of the second type of candidate users under the user characteristic dimension are generated, and a preference level label with a low degree of preference is added to the second type of candidate users, Get sample data of negative sample users.

In an embodiment of the present application, when generating sample data of negative sample users, multiple user feature vectors of the second-type candidate users in the user-feature dimension may be generated based on the user feature information of the second-type candidate users, and Add preference grade labels with low preference for the second type of candidate users.

Optionally, for the second type of candidate users, all candidate users in the second type of candidate users may be selected as negative sample users.

Referring to FIG. 9 , FIG. 9 shows a specific flow chart of step S740 of a method for generating a user portrait according to an embodiment of the present application. The step S740 may include steps S910 to S920 , which are described in detail as follows.

In step S910, candidate users are extracted from the second type of candidate users.

Referring to FIG. 10 , FIG. 10 shows a specific flowchart of step S910 of a method for generating a user portrait according to an embodiment of the present application. The step S910 may include steps S1010 to S1020 , which are described in detail as follows.

In step S1010, based on the frequency of the historical behavior data included in the user characteristic information of the second-type candidate user, the portrait density of the second-type candidate user is determined, and the portrait density and the frequency are negatively correlated.

In an embodiment of the present application, for the second type of candidate users, all candidate users in the second type of candidate users may be selected as negative sample users.

In an embodiment of the present application, since the second type of candidate users is a candidate user who is not in the target operation event in the historical behavior data, there are two types of candidate users with rich historical behavior and sparse historical behavior. When the historical behavior of the candidate user is rich And when the historical behavior data is not in the target operation event, the confidence of the candidate user as a negative sample user is high. When the historical behavior of the candidate user is sparse and the historical behavior data is not in the target operation event, there will be some candidate users. It is the case where users are selected as negative samples due to sparse historical behavior. Therefore, it is necessary to select candidate samples with relatively more historical behaviors as negative sample users, and select candidate samples with relatively few historical behaviors as negative sample users.

Portrait density is a measure that reflects the richness of user historical behavior. The greater the portrait density, the richer the user's historical behavior.

The portrait density can be determined according to the frequency of the historical behavior data included in the user characteristic information of the second type of candidate users, and the portrait density is positively correlated with the frequency of the historical behavior data. The frequency of historical behavioral data can be calculated in a number of ways.

Optionally, the sum of the frequencies of various types of operation events included in the historical behavior data may be used as the frequency of the historical behavior data.

Optionally, different weights may be assigned to various types of operation events included in the historical behavior data, and the weighted sum of the frequencies of the various types of operation events included in the historical behavior data may be used as the frequency of the historical behavior data.

In step S1020, among the second type of candidate users whose portrait density is higher than a predetermined portrait density threshold, the candidate users are extracted at a first ratio, and among the second type of candidate users whose portrait density is lower than or equal to the predetermined portrait density threshold Among the class candidate users, candidate users are extracted in a second proportion, and the first proportion is greater than the second proportion.

In one embodiment of the present application, the profile density threshold is a metric to determine whether the candidate user is a user with rich historical behavior. If the profile density is higher than a predetermined profile density threshold, the candidate user is a user with rich historical behavior. , if the profile density is lower than or equal to the predetermined profile density threshold, the candidate user is a user with sparse historical behavior. Among the second type of candidate users whose portrait density is higher than the predetermined portrait density threshold, the candidate users can be extracted with a first ratio, and the second type of candidate users whose portrait density is lower than or equal to the predetermined portrait density threshold , candidate users can be extracted at a second ratio, and the first ratio can be set to be greater than the second ratio

In the technical solution of the embodiment shown in FIG. 10 , the candidate samples with relatively more historical behaviors are selected as negative sample users by selecting the candidate samples with relatively more historical behaviors as the negative sample users, and the historical behaviors relatively few users are selected in the historical behavior data that are not among the second-type candidate users of the target operation event. As a negative sample user, the selected candidate sample can improve the accuracy of the selected negative sample user, thereby improving the classification confidence of the pre-trained machine learning model.

Please also continue to refer to Fig. 9, in step S920, based on the user feature information of the extracted candidate user, generate multiple user feature vectors of the extracted candidate user under the user feature dimension, and add a low preference level for the extracted candidate user. Preference level label to get the sample data of negative sample users.

In an embodiment of the present application, for the candidate users extracted from the second type of candidate users, when generating the corresponding sample data, the extracted candidate users can be generated based on the user feature information of the extracted candidate users. Multiple user feature vectors under the feature dimension are added, and preference grade labels with low preference are added to the extracted candidate users, thereby generating sample data of negative sample users.

Please continue to refer to FIG. 7, in step S750, based on the sample data of the positive sample users and the sample data of the negative sample users, the training set sample data for training the machine learning model to be trained is obtained.

In one embodiment, after obtaining the sample data of the positive sample users and the sample data of the negative sample users, the obtained sample data of the positive sample users and the sample data of the negative sample users can be used as the training set sample data to Realize the training of the machine learning model to be trained through the sample data of the training set.

Please continue to refer to FIG. 6 , in step S620 , the machine learning model to be trained is trained by using the sample data of the training set to obtain a pre-trained machine learning model.

In one embodiment of the present application, the sample data of the training set is input into the machine learning model to be trained, and the machine learning model to be trained is trained by the sample data of the training set to obtain a pre-trained machine learning model. The process of training the machine learning model is to adjust the coefficients in the network layer corresponding to the machine learning model, so that the input target user's multiple user feature vectors under the user feature dimension can pass through the network layer corresponding to the machine learning model. The coefficient operation of , and the output obtains the preference level label of the target user for the target object under the target classification attribute.

Please also continue to refer to FIG. 2 , in step S240 , the preference level label of the target user for the target object under the target classification attribute output by the pre-trained machine learning model is obtained.

In an embodiment of the present application, the preference level label of the target user for the target object under the target classification attribute output by the pre-trained machine learning model can be obtained, so as to obtain the target user's preference for the target object under the target classification attribute.

In step S250, if the acquired preference level label matches the preset preference level label, a user portrait of the target user is generated based on the target classification attribute.

In an embodiment of the present application, the preset preference level label is a preference level label representing a higher degree of preference of the target user to the target attribute of the target classification attribute, such as the preference level of “interested” and “specially interested” tags, or just "particularly interesting" preference rating tags. When the preference level label obtained from the pre-trained machine learning model matches the preset preference level label, it can be determined that the target user has a high preference level for the target object under the target classification attribute, so it can be generated according to the target classification attribute. User persona of the target user.

It can be seen from the above that the user characteristic information of the target user is obtained, and the user characteristic information includes the user attribute information and historical behavior data of the target user; based on the user characteristic information, multiple user characteristic vectors of the target user under the user characteristic dimension are respectively generated. The feature dimension refers to the feature dimension composed of the category of user attribute information, the operation event category contained in the historical behavior data, and the classification attribute of the operation object; multiple user feature vectors are input into the pre-training machine learning model, and the pre-training The machine learning model contains multiple user feature vectors of the sample user under the user feature dimension and the sample user's preference level label for the target object under the target classification attribute; obtain the target user's target classification output by the pre-trained machine learning model The preference grade label of the target object under the attribute; if the obtained preference grade label matches the preset preference grade label, a user portrait of the target user is generated based on the target classification attribute. Compared with obtaining user portraits based on tag statistics, user feature vectors can more comprehensively represent user preferences, which can improve the accuracy of predicting users' preference grade labels for target objects under specific classification attributes, thereby facilitating the identification of specific users. The target users of the portraits, in order to improve the accuracy of related business recommendations.

The apparatus embodiments of the present application are described below, which can be used to execute the method for generating a user portrait in the above-mentioned embodiments of the present application. For details not disclosed in the device embodiments of the present application, please refer to the above-mentioned method embodiments of the present application.

FIG. 11 shows a block diagram of an apparatus for generating a user portrait according to an embodiment of the present application.

Referring to FIG. 11 , an apparatus 1100 according to an embodiment of the present application includes: a first obtaining unit 1110 , a first generating unit 1120 , an input unit 1130 , an output unit 1140 and a second generating unit 1150 . Wherein, the first obtaining unit 1110 is used to obtain user characteristic information of the target user, the user characteristic information includes the user attribute information and historical behavior data of the target user; the first generating unit 1120 is used to obtain the user characteristic information based on the user characteristic information, and respectively generate multiple user feature vectors of the target user under the user feature dimension, where the user feature dimension refers to the category of the user attribute information and the operation event category and operation included in the historical behavior data. The feature dimension formed by the classification attributes of the object; the input unit 1130 is used for inputting the plurality of user feature vectors into a pre-trained machine learning model, and the pre-trained machine learning model contains sample users in the user feature dimension A plurality of user feature vectors and the preference grade labels of the sample users to the target object under the target classification attribute; output unit 1140 is used to obtain the target user output of the pre-trained machine learning model to the target object under the target classification attribute. The second generation unit 1150 is configured to generate the user portrait of the target user based on the target classification attribute if the acquired preference grade label matches the preset preference grade label.

In some embodiments of the present application, based on the foregoing solution, the device for generating user portraits further includes: a third generating unit, configured to perform conversion processing on the plurality of user feature vectors respectively to generate the plurality of user features The first feature vector corresponding to the vector, the first feature vector corresponding to the multiple user features is a vector of the same dimension; the aggregation unit is used for the preset association relationship between the multiple user feature information. The first feature vectors corresponding to the plurality of user feature vectors are aggregated to generate a plurality of second feature vectors after aggregation; a prediction unit is used to predict the target user's response to the target based on the plurality of second feature vectors after the aggregation. The preference level label of the target object under the classification attribute, and the first feature vector corresponding to the multiple user features is a vector of the same dimension.

In some embodiments of the present application, based on the foregoing solution, the device for generating user portraits further includes: a second acquiring unit, configured to acquire training set sample data for training the machine learning model to be trained, and the training Each piece of sample data in the set sample data includes multiple user feature vectors of the sample user under the user feature dimension and the sample user's preference level label for the target object under the target classification attribute; the training unit is used to pass the training set sample The data is used to train the machine learning model to be trained to obtain a pre-trained machine learning model.

In some embodiments of the present application, based on the foregoing solution, the second obtaining unit is configured to: obtain user feature information of candidate users in the candidate user set; for each candidate user in the candidate user set, based on the candidate user The historical behavior data in the user feature information of the A candidate user of an operation event, the target operation event refers to an operation event that operates the target object under the target classification attribute; based on the user feature information of the first type of candidate user, the first type of candidate user is generated Multiple user feature vectors under the user feature dimension, and add preference level labels with high preference to the first type of candidate users to obtain sample data of positive sample users; based on the user characteristic information of the second type of candidate users , generate a plurality of user feature vectors of the second type of candidate users under the user characteristic dimension, and add a preference level label with a low degree of preference to the second type of candidate users to obtain sample data of negative sample users; based on the The sample data of the positive sample user and the sample data of the negative sample user are used to obtain the training set sample data for training the machine learning model to be trained.

In some embodiments of the present application, based on the foregoing solution, the second obtaining unit is configured to: extract candidate users from the second type of candidate users; generate the extracted candidate users based on user feature information of the extracted candidate users Multiple user feature vectors of the candidate users under the user feature dimension, and adding a preference level label with a low degree of preference to the extracted candidate users to obtain sample data of negative sample users.

In some embodiments of the present application, based on the foregoing solution, the second acquisition unit is configured to: determine the second type of candidate user based on the frequency of historical behavior data included in the user feature information of the second type of candidate user The portrait density is positively correlated with the frequency; in the second type of candidate users whose portrait density is higher than the predetermined portrait density threshold, the candidate users are extracted at the first ratio, and the Among the candidate users of the second category whose portrait density is lower than or equal to a predetermined portrait density threshold, candidate users are extracted with a second proportion, and the first proportion is greater than the second proportion.

In some embodiments of the present application, based on the foregoing solution, the second obtaining unit is configured to: obtain a configuration file, where the configuration file is used to identify the attribute category of the obtained user attribute information and the obtained historical behavior data. The operation event category and the classification attribute of the operation object are configured; for each candidate user in the candidate user set, based on the category of the user attribute information configured in the configuration file, the user of the candidate user under the configured attribute category is obtained. Attribute information; for each candidate user in the candidate user set, based on the operation event category and the classification attribute of the operation object configured in the configuration file, obtain the candidate user's information under the operation event category and the classification attribute of the operation object. Historical behavioral data.

FIG. 12 shows a schematic structural diagram of a computer system suitable for implementing the electronic device according to the embodiment of the present application.

It should be noted that the computer system 1200 of the electronic device shown in FIG. 12 is only an example, and should not impose any limitations on the functions and scope of use of the embodiments of the present application.

As shown in FIG. 12 , the computer system 1200 includes a central processing unit (Central Processing Unit, CPU) 1201, which can be loaded into a random device according to a program stored in a read-only memory (Read-Only Memory, ROM) 1202 or from a storage part 1208 A program in a random access memory (RAM) 1203 is accessed to perform various appropriate actions and processes, for example, the methods described in the above embodiments are performed. In the RAM 1203, various programs and data necessary for system operation are also stored. The CPU 1201 , the ROM 1202 , and the RAM 1203 are connected to each other through a bus 1204 . An Input/Output (I/O) interface 1205 is also connected to the bus 1204 .

The following components are connected to the I/O interface 1205: an input section 1206 including a keyboard, a mouse, etc.; an output section 1207 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc. ; a storage part 1208 including a hard disk and the like; and a communication part 1209 including a network interface card such as a LAN (Local Area Network) card, a modem, and the like. The communication section 1209 performs communication processing via a network such as the Internet. Drivers 1210 are also connected to I/O interface 1205 as needed. A removable medium 1211, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is mounted on the drive 1210 as needed so that a computer program read therefrom is installed into the storage section 1208 as needed.

In particular, according to embodiments of the present application, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, embodiments of the present application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program comprising a computer program for performing the method illustrated in the flowchart. In such an embodiment, the computer program may be downloaded and installed from the network via the communication portion 1209, and/or installed from the removable medium 1211. When the computer program is executed by the central processing unit (CPU) 1201, various functions defined in the system of the present application are executed.

It should be noted that the computer-readable medium shown in the embodiments of the present application may be a computer-readable signal medium or a computer-readable storage medium, or any combination of the above two. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus or device, or a combination of any of the above. More specific examples of computer readable storage media may include, but are not limited to, electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read only memory (ROM), erasable Erasable Programmable Read Only Memory (EPROM), flash memory, optical fiber, portable Compact Disc Read-Only Memory (CD-ROM), optical storage device, magnetic storage device, or any suitable of the above The combination. In this application, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this application, however, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying a computer-readable computer program therein. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing. A computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device . A computer program embodied on a computer-readable medium may be transmitted using any suitable medium, including but not limited to: wireless, wired, etc., or any suitable combination of the foregoing.

The flowchart and block diagrams in the Figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. Wherein, each block in the flowchart or block diagram may represent a module, program segment, or part of code, and the above-mentioned module, program segment, or part of code contains one or more executables for realizing the specified logical function instruction. It should also be noted that, in some alternative implementations, the functions noted in the blocks may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It is also noted that each block of the block diagrams or flowchart illustrations, and combinations of blocks in the block diagrams or flowchart illustrations, can be implemented in special purpose hardware-based systems that perform the specified functions or operations, or can be implemented using A combination of dedicated hardware and computer instructions is implemented.

The units involved in the embodiments of the present application may be implemented in software or hardware, and the described units may also be provided in a processor. Among them, the names of these units do not constitute a limitation on the unit itself under certain circumstances.

As another aspect, the present application also provides a computer-readable medium. The computer-readable medium may be included in the electronic device described in the above embodiments; it may also exist alone without being assembled into the electronic device. middle. The above-mentioned computer-readable medium carries one or more programs, and when the above-mentioned one or more programs are executed by an electronic device, enables the electronic device to implement the methods described in the above-mentioned embodiments.

It should be noted that although several modules or units of the apparatus for action performance are mentioned in the above detailed description, this division is not mandatory. Indeed, according to embodiments of the present application, the features and functions of two or more modules or units described above may be embodied in one module or unit. Conversely, the features and functions of one module or unit described above may be further divided into multiple modules or units to be embodied.

From the description of the above embodiments, those skilled in the art can easily understand that the exemplary embodiments described herein may be implemented by software, or may be implemented by software combined with necessary hardware. Therefore, the technical solutions according to the embodiments of the present application may be embodied in the form of software products, and the software products may be stored in a non-volatile storage medium (which may be CD-ROM, U disk, mobile hard disk, etc.) or on the network , which includes several instructions to cause a computing device (which may be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the embodiments of the present application.

Other embodiments of the present application will readily occur to those skilled in the art upon consideration of the specification and practice of the embodiments disclosed herein. This application is intended to cover any variations, uses or adaptations of this application that follow the general principles of this application and include common knowledge or conventional techniques in the technical field not disclosed in this application .

It is to be understood that the present application is not limited to the precise structures described above and illustrated in the accompanying drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the application is limited only by the appended claims.

Claims (10)

1. A method for generating a user representation, comprising:

acquiring user characteristic information of a target user, wherein the user characteristic information comprises user attribute information and historical behavior data of the target user;

respectively generating a plurality of user feature vectors of the target user under a user feature dimension based on the user feature information, wherein the user feature dimension is a feature dimension formed by the category of the user attribute information, the operation event category contained in the historical behavior data and the classification attribute of the operation object;

inputting the plurality of user feature vectors into a pre-trained machine learning model, wherein the pre-trained machine learning model comprises a plurality of user feature vectors of sample users under user feature dimensions and preference level labels of the sample users for target objects under target classification attributes;

acquiring a preference level label of a target object under the target classification attribute of a target user output by the pre-trained machine learning model;

and if the acquired preference level label is matched with a preset preference level label, generating a user portrait of the target user based on the target classification attribute.

2. The method of generating a user representation as claimed in claim 1 wherein the pre-trained machine learning model determines a preference level label of a target user for a target object under a target classification attribute based on:

respectively carrying out conversion processing on the plurality of user characteristic vectors to generate first characteristic vectors corresponding to the plurality of user characteristic vectors, wherein the first characteristic vectors corresponding to the plurality of user characteristics are vectors with the same dimensionality;

based on preset incidence relations among the plurality of user feature information, carrying out aggregation processing on first feature vectors corresponding to the plurality of user feature vectors to generate a plurality of aggregated second feature vectors, wherein the first feature vectors corresponding to the plurality of user features are vectors with the same dimensionality;

and predicting preference level labels of the target users on the target objects under the target classification attributes based on the aggregated second feature vectors.

3. A user representation generation method as claimed in claim 1, further comprising:

acquiring training set sample data for training a machine learning model to be trained, wherein each piece of sample data in the training set sample data comprises a plurality of user feature vectors of a sample user under a user feature dimension and a preference level label of the sample user for a target object under a target classification attribute;

and training the machine learning model to be trained through the training set sample data to obtain the pre-trained machine learning model.

4. The method of generating a user representation of claim 3, wherein said obtaining training set sample data for training a machine learning model to be trained comprises:

acquiring user characteristic information of candidate users in a candidate user set;

for each candidate user in the candidate user set, determining a first class of candidate users and a second class of candidate users based on historical behavior data in user feature information of the candidate users, wherein the first class of candidate users are candidate users including target operation events, the second class of candidate users are candidate users not including the target operation events, and the target operation events refer to operation events for operating target objects under the target classification attributes;

generating a plurality of user characteristic vectors of the first type candidate users under the user characteristic dimension based on the user characteristic information of the first type candidate users, and adding preference grade labels with high preference degrees to the first type candidate users to obtain sample data of a positive sample user;

generating a plurality of user feature vectors of the second type candidate users under the user feature dimension based on the user feature information of the second type candidate users, and adding preference level labels with low preference degrees to the second type candidate users to obtain sample data of negative sample users;

and obtaining training set sample data used for training a machine learning model to be trained based on the sample data of the positive sample user and the sample data of the negative sample user.

5. The method of claim 4, wherein the generating a plurality of user feature vectors of the second type candidate users in the user feature dimension based on the user feature information of the second type candidate users and adding preference level labels with low preference degrees to the second type candidate users to obtain sample data of negative sample users comprises:

extracting candidate users from the second type of candidate users;

and generating a plurality of user feature vectors of the extracted candidate users under the user feature dimension based on the extracted user feature information of the candidate users, and adding preference grade labels with low preference degrees to the extracted candidate users to obtain sample data of the negative sample users.

6. The method of claim 5, wherein the extracting candidate users from the second category of candidate users comprises:

determining portrait intensity of the second type candidate users based on frequency of historical behavior data contained in user characteristic information of the second type candidate users, wherein the portrait intensity and the frequency are in positive correlation;

candidate users are extracted at a first ratio among the second category of candidate users having a portrait intensity higher than a predetermined portrait intensity threshold, and candidate users are extracted at a second ratio among the second category of candidate users having a portrait intensity lower than or equal to the predetermined portrait intensity threshold, the first ratio being greater than the second ratio.

7. The method of claim 4, wherein the obtaining user feature information of the candidate users in the candidate user set comprises:

acquiring a configuration file, wherein the configuration file is used for configuring the attribute category of the acquired user attribute information, the operation event category of the acquired historical behavior data and the classification attribute of an operation object;

for each candidate user in the candidate user set, acquiring user attribute information of the candidate user under the configured attribute category based on the category of the user attribute information configured in the configuration file;

and for each candidate user in the candidate user set, acquiring historical behavior data of the candidate user under the operation event category and the classification attribute of the operation object based on the operation event category and the classification attribute of the operation object configured in the configuration file.

8. An apparatus for generating a user representation, comprising:

the device comprises a first acquisition unit, a second acquisition unit and a third acquisition unit, wherein the first acquisition unit is used for acquiring user characteristic information of a target user, and the user characteristic information comprises user attribute information and historical behavior data of the target user;

a first generating unit, configured to generate, based on the user feature information, a plurality of user feature vectors of the target user in a user feature dimension, where the user feature dimension is a feature dimension formed by a category of the user attribute information, and an operation event category and a classification attribute of an operation object included in the historical behavior data;

the input unit is used for inputting the plurality of user feature vectors into a pre-trained machine learning model, and the pre-trained machine learning model comprises a plurality of user feature vectors of sample users under user feature dimensions and preference level labels of the sample users for target objects under target classification attributes;

the output unit is used for acquiring a preference level label of the target user on the target object under the target classification attribute output by the pre-trained machine learning model;

and the second generation unit is used for generating the user portrait of the target user based on the target classification attribute if the acquired preference level label is matched with a preset preference level label.

9. An electronic device, comprising:

one or more processors;

storage means for storing one or more programs which, when executed by the one or more processors, cause the one or more processors to carry out a method of user representation generation as claimed in any one of claims 1 to 7.

10. A computer program medium having computer readable instructions stored thereon which, when executed by a processor, implement a method of user representation generation as claimed in any of claims 1 to 7.

Images