Data processing methods, apparatus, electronic devices and storage media

Abstract

This disclosure provides a data processing method, apparatus, electronic device, and storage medium. This disclosure relates to the field of computer technology, and particularly to the fields of artificial intelligence, data mining, and machine learning. The implementation scheme includes: acquiring a sample dataset for modeling; dividing a first number of feature dimensions into multiple categories; selecting multiple feature dimensions from each of the multiple categories to obtain a second number of feature dimensions; and performing dimensionality reduction processing on each sample data in the sample dataset, including: deleting feature dimensions other than the second number of feature dimensions; and for each feature dimension in the second number of feature dimensions with a positive feature value, replacing the feature value of that feature dimension with a semantic vector corresponding to that feature dimension.

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and more particularly to the fields of artificial intelligence, data mining and machine learning, specifically to a data processing method, apparatus, electronic device, computer-readable storage medium and computer program product. Background Technology

[0002] Machine learning is the science of artificial intelligence. Its main research focus is on artificial intelligence, particularly how to improve the performance of specific algorithms through experiential learning. It is primarily divided into supervised learning, unsupervised learning, and reinforcement learning. Among these, supervised learning involves scenarios where modeling is based on structured, high-dimensional, sparse sample datasets.

[0003] The methods described in this section are not necessarily methods that had been previously conceived or adopted. Unless otherwise specified, no method described in this section should be assumed to be prior art simply because it is included in this section. Similarly, unless otherwise specified, the issues mentioned in this section should not be considered to be accepted in any prior art. Summary of the Invention

[0004] This disclosure provides a data processing method, apparatus, electronic device, computer-readable storage medium, and computer program product.

[0005] According to one aspect of this disclosure, a data processing method is provided, comprising: acquiring a sample dataset for modeling, wherein each sample data in the sample dataset is high-dimensional sparse data and includes a first number of feature dimensions; dividing the first number of feature dimensions into multiple categories; selecting multiple feature dimensions from each of the multiple categories to obtain a second number of feature dimensions; and performing dimensionality reduction processing on each sample data in the sample dataset, the dimensionality reduction processing including: deleting feature dimensions other than the second number of feature dimensions; and for each feature dimension in the second number of feature dimensions whose feature values ​​are positive, replacing the feature value of the feature dimension with a semantic vector corresponding to the feature dimension, wherein the semantic vector is obtained through unsupervised training.

[0006] According to another aspect of this disclosure, a model training method is provided, comprising: acquiring a sample dataset, wherein each sample data in the sample dataset is obtained based on the data processing method described above and includes a corresponding sample label; initializing multiple parameters of the model; and for each sample data, performing the following operations: inputting the sample data into the model to obtain a model calculation result; and adjusting multiple parameters of the model based on the model calculation result and the corresponding sample label of the sample data.

[0007] According to another aspect of this disclosure, a data processing apparatus is provided, comprising: a first acquisition unit configured to acquire a sample dataset for modeling, wherein each sample data in the sample dataset is high-dimensional sparse data and includes a first number of feature dimensions; a partitioning unit configured to partition the first number of feature dimensions into multiple categories; a second acquisition unit configured to select multiple feature dimensions from each of the multiple categories to acquire a second number of feature dimensions; and a dimensionality reduction processing unit configured to perform dimensionality reduction processing on each sample data in the sample dataset, the dimensionality reduction processing unit comprising: a deletion subunit configured to delete feature dimensions other than the second number of feature dimensions; and a replacement subunit configured to, for each feature dimension in the second number of feature dimensions whose feature values ​​are positive, replace the feature value of the feature dimension with a semantic vector corresponding to the feature dimension, wherein the semantic vector is obtained through unsupervised training.

[0008] According to another aspect of this disclosure, a model training apparatus is provided, comprising: a third acquisition unit configured to acquire a sample dataset, wherein each sample data in the sample dataset is obtained based on the data processing method described above and includes a corresponding sample label; an initialization unit configured to initialize multiple parameters of the model; and an execution unit configured to perform the operations of the following subunits for each sample data, the execution unit comprising: an input subunit configured to input the sample data into the model to obtain a model calculation result; and an adjustment subunit configured to adjust multiple parameters of the model based on the model calculation result and the corresponding sample label of the sample data.

[0009] According to another aspect of this disclosure, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the aforementioned data processing method or model training method.

[0010] According to another aspect of this disclosure, a non-transitory computer-readable storage medium is provided that stores computer instructions, wherein the computer instructions are used to cause a computer to perform the above-described data processing method or model training method.

[0011] According to another aspect of this disclosure, a computer program product is provided, including a computer program, wherein the computer program implements the above-described data processing method or model training method when executed by a processor.

[0012] According to one or more embodiments of this disclosure, the dimensionality of sample data can be reduced while ensuring the model training effect, thereby reducing the model computation and improving the model training efficiency.

[0013] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0014] The accompanying drawings exemplify embodiments and form part of the specification, serving together with the textual description to explain exemplary implementations of the embodiments. The illustrated embodiments are for illustrative purposes only and do not limit the scope of the claims. Throughout the drawings, the same reference numerals refer to similar but not necessarily identical elements.

[0015] Figure 1 A schematic diagram of an exemplary system in which the various methods described herein may be implemented according to embodiments of the present disclosure is shown;

[0016] Figure 2 A flowchart of a data processing method according to an embodiment of the present disclosure is shown;

[0017] Figure 3 A flowchart of a data processing method according to an embodiment of the present disclosure is shown;

[0018] Figure 4 A schematic diagram of an unsupervised semantic representation model according to an embodiment of the present disclosure is shown;

[0019] Figure 5 A flowchart of a model training method according to an embodiment of the present disclosure is shown;

[0020] Figure 6 A structural block diagram of a data processing apparatus according to an embodiment of the present disclosure is shown;

[0021] Figure 7 A structural block diagram of a data processing apparatus according to an embodiment of the present disclosure is shown;

[0022] Figure 8 A structural block diagram of a model training apparatus according to an embodiment of the present disclosure is shown;

[0023] Figure 9 A structural block diagram of an exemplary electronic device that can be used to implement embodiments of the present disclosure is shown. Detailed Implementation

[0024] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0025] In this disclosure, unless otherwise stated, the use of terms such as "first," "second," etc., to describe various elements is not intended to limit the positional, temporal, or importance relationships of these elements; such terms are merely used to distinguish one element from another. In some examples, the first element and the second element may refer to the same instance of that element, while in other cases, based on the context, they may refer to different instances.

[0026] The terminology used in the description of the various examples described in this disclosure is for the purpose of describing particular examples only and is not intended to be limiting. Unless the context explicitly indicates otherwise, an element may be one or more unless the number of elements is specifically limited. Furthermore, the term "and / or" as used in this disclosure covers any one of the listed items and all possible combinations thereof.

[0027] The embodiments of this disclosure will now be described in detail with reference to the accompanying drawings.

[0028] Figure 1 A schematic diagram of an exemplary system 100 in which the various methods and apparatus described herein can be implemented according to embodiments of this disclosure is shown. Reference Figure 1 The system 100 includes one or more client devices 101, 102, 103, 104, 105 and 106, a server 120, and one or more communication networks 110 coupling the one or more client devices to the server 120. The client devices 101, 102, 103, 104, 105 and 106 can be configured to execute one or more applications.

[0029] In embodiments of this disclosure, server 120 may run one or more services or software applications that enable the execution of data processing methods.

[0030] In some embodiments, server 120 may also provide other services or software applications that may include non-virtual and virtual environments. In some embodiments, these services may be provided as web-based services or cloud services, such as to users of client devices 101, 102, 103, 104, 105 and / or 106 under a Software as a Service (SaaS) model.

[0031] exist Figure 1 In the configuration shown, server 120 may include one or more components that implement the functions performed by server 120. These components may include software components, hardware components, or combinations thereof that can be executed by one or more processors. Users operating client devices 101, 102, 103, 104, 105, and / or 106 can sequentially interact with server 120 using one or more client applications to utilize the services provided by these components. It should be understood that various different system configurations are possible and may differ from system 100. Therefore, Figure 1 This is an example of a system used to implement the various methods described herein, and is not intended to be limiting.

[0032] Users can upload sample data using client devices 101, 102, 103, 104, 105, and / or 106. The client devices can provide an interface that allows users to interact with them. The client devices can also output information to users through this interface. Although... Figure 1 Only six client devices are described, but those skilled in the art will understand that this disclosure can support any number of client devices.

[0033] Client devices 101, 102, 103, 104, 105, and / or 106 may include various types of computer devices, such as portable handheld devices, general-purpose computers (such as personal computers and laptops), workstation computers, wearable devices, smart screen devices, self-service terminal devices, service robots, gaming systems, thin clients, various messaging devices, sensors, or other sensing devices. These computer devices can run various types and versions of software applications and operating systems, such as Microsoft Windows, Apple iOS, UNIX-like operating systems, Linux or Linux-like operating systems (such as Google Chrome OS); or include various mobile operating systems, such as Microsoft Windows Mobile OS, iOS, Windows Phone, and Android. Portable handheld devices may include cellular phones, smartphones, tablets, personal digital assistants (PDAs), etc. Wearable devices may include head-mounted displays (such as smart glasses) and other devices. Gaming systems may include various handheld gaming devices, internet-enabled gaming devices, etc. Client devices are capable of executing various applications, such as various internet-related applications, communication applications (such as email applications), short message service (SMS) applications, and can use various communication protocols.

[0034] Network 110 can be any type of network well known to those skilled in the art, and can use any of a variety of available protocols (including but not limited to TCP / IP, SNA, IPX, etc.) to support data communication. By way of example only, one or more networks 110 can be a local area network (LAN), an Ethernet-based network, a token ring network, a wide area network (WAN), the Internet, a virtual network, a virtual private network (VPN), an intranet, an extranet, a public switched telephone network (PSTN), an infrared network, a wireless network (e.g., Bluetooth, WIFI), and / or any combination of these and / or other networks.

[0035] Server 120 may include one or more general-purpose computers, special-purpose server computers (e.g., PC (personal computer) servers, UNIX servers, mid-range servers), blade servers, mainframe computers, server clusters, or any other suitable arrangement and / or combination. Server 120 may include one or more virtual machines running a virtual operating system, or other computing architectures involving virtualization (e.g., one or more flexible pools of logical storage devices that can be virtualized to maintain virtual storage devices for servers). In various embodiments, server 120 may run one or more services or software applications that provide the functionality described below.

[0036] The computing unit in server 120 can run one or more operating systems, including any of the aforementioned operating systems and any commercially available server operating system. Server 120 can also run any of a variety of additional server applications and / or middleware applications, including HTTP servers, FTP servers, CGI servers, JAVA servers, database servers, etc.

[0037] In some implementations, server 120 may include one or more applications to analyze and merge data feeds and / or event updates received from users of client devices 101, 102, 103, 104, 105, and 106. Server 120 may also include one or more applications to display data feeds and / or real-time events via one or more display devices of client devices 101, 102, 103, 104, 105, and 106.

[0038] In some implementations, server 120 can be a server for a distributed system or a server integrated with blockchain. Server 120 can also be a cloud server, or an intelligent cloud computing server or intelligent cloud host with artificial intelligence technology. A cloud server is a host product in the cloud computing service system, designed to address the shortcomings of traditional physical hosts and Virtual Private Server (VPS) services, such as high management difficulty and weak business scalability.

[0039] System 100 may also include one or more databases 130. In some embodiments, these databases may be used to store data and other information. For example, one or more of the databases 130 may be used to store information such as audio files and video files. The data repository 130 may reside in various locations. For example, a data repository used by server 120 may be local to server 120, or it may be located remotely to server 120 and may communicate with server 120 via a network-based or dedicated connection. The data repository 130 may be of different types. In some embodiments, the data repository used by server 120 may be a database, such as a relational database. One or more of these databases may store, update, and retrieve data from and from the database in response to commands.

[0040] In some embodiments, one or more of the databases 130 may also be used by an application to store application data. The databases used by the application may be of different types, such as key-value stores, object stores, or regular stores supported by a file system.

[0041] Figure 1 The system 100 can be configured and operated in various ways to enable the application of the various methods and apparatus described in this disclosure.

[0042] In the field of machine learning, there are often many scenarios where the sample data used to train models is high-dimensional and sparse, such as user behavior modeling. High-dimensional sparse data often has hundreds of thousands or even millions of dimensions, but only a few hundred dimensions are typically representative of features. For example, user behavior data is often high-dimensional and sparse. Due to the high-dimensional and sparse nature of the sample data, using such sample data for model training and inference often results in high computational costs and low training or inference efficiency. Furthermore, models trained on high-dimensional and sparse sample datasets often exhibit poor generalization performance; even slight fluctuations in the input data can lead to significant differences in the output results.

[0043] According to embodiments of this disclosure, such as Figure 2As shown, a data processing method 200 is provided, including: step S201, obtaining a sample dataset for modeling, wherein each sample data in the sample dataset is high-dimensional sparse data and includes a first number of feature dimensions; step S202, dividing the first number of feature dimensions into multiple categories; step S203, selecting multiple feature dimensions from each of the multiple categories to obtain a second number of feature dimensions; and step S204, performing dimensionality reduction processing on each sample data in the sample dataset, the dimensionality reduction processing including: step S205, deleting feature dimensions other than the second number of feature dimensions; and step S206, for each feature dimension in the second number of feature dimensions with a positive feature value, replacing the feature value of the feature dimension with a semantic vector corresponding to the feature dimension.

[0044] Based on the embodiments of this disclosure, the dimensionality of the sample data is reduced by reconstructing the high-dimensional sparse sample data according to a preset category, thereby solving the problems of large modeling computation and low training or inference efficiency of the model; at the same time, the feature dimensions with positive feature values ​​are replaced by corresponding semantic vectors, and the obtained sample dataset is used for model training, thereby solving the problem of poor model generalization ability.

[0045] In some supervised learning scenarios, the sample data used for modeling can be high-dimensional sparse data. For example, in user behavior modeling, the sample dataset is user behavior sample data, which is a type of high-dimensional sparse data. It is usually a fixed-length structured vector, where each position of the vector represents a feature dimension of the user behavior feature, and each feature dimension has a feature value that represents a specific meaning.

[0046] In one example, a positive value for a feature dimension indicates that the user possesses that feature; conversely, a negative value for a feature dimension indicates that the user does not possess that feature. For instance, a feature dimension with a value of 1 indicates that the user possesses that feature; conversely, a feature dimension with a value of 0 or -1 indicates that the user does not possess that feature. It is understood that the "1" in the above example is merely an example of a positive value in this disclosure; other examples are conceived and this disclosure does not limit the scope of the expression.

[0047] A sample of data can have a maximum number of feature dimensions. Taking user behavior data as an example, a sample of user behavior data can have 300,000 feature dimensions. However, due to the significant differences in various feature dimensions such as user interest attributes, most users have very few features. Generally, there may only be a few hundred feature dimensions with positive values.

[0048] The first number of feature dimensions can be divided into multiple categories based on the actual situation.

[0049] In one example, the sample data could be behavioral data about users installing mobile applications. The sample data could be categorized based on the features of the mobile applications. For example, multiple categories could include system applications, entertainment and health, games and card games, lifestyle and shopping, finance and current affairs, etc.

[0050] In one example, the sample data could also be user interest data. Based on the content of the interests, the feature dimensions of the sample data can be classified. For example, multiple categories could be sports and e-sports, digital electronics, maternal and infant education, finance and economics, current affairs and politics, etc.

[0051] It is understandable that feature dimensions can be divided according to actual circumstances and needs. The above example is only an example of categorization based on function or content. Feature dimension categories can also be divided based on other criteria, which are not limited here.

[0052] In some embodiments, selecting multiple feature dimensions from each of the multiple categories to obtain a second number of feature dimensions may include: for each of the multiple categories, dividing all feature dimensions included in that category into multiple subcategories; and selecting multiple feature dimensions from each of the multiple subcategories to constitute the second number of feature dimensions. This allows for a more balanced acquisition of features from each category when performing dimensionality reduction on sample data, avoiding the problem of unbalanced feature distribution in the dimensionality-reduced sample data.

[0053] For each category, all feature dimensions can be further divided into multiple subcategories.

[0054] In one example, the sample data could be behavioral data about users installing mobile applications. The sample data could be categorized and subcategorized based on the features of the mobile applications. For example, for entertainment and health-related mobile applications, they could be further divided into multiple subcategories such as health and medical care, map and travel, audio and video playback, photography and videography, and lifestyle and leisure. Similarly, for system application mobile applications, they could be further divided into multiple subcategories such as office and business, mobile phone customization, system security, and auxiliary tools.

[0055] In one example, the feature dimensions of the sample data can also be classified based on the content provided by the mobile application. For example, for mobile applications of games and card games, they can be further divided into multiple subcategories such as action-adventure, role-playing, poker and card games, and management strategy based on the game content they provide.

[0056] Understandably, feature dimensions can be subdivided according to actual circumstances and needs. The above example is only an example of subdividing based on function or content. Feature dimension subdivision can also be based on other criteria, which are not limited here.

[0057] In some embodiments, the number of feature dimension categories and subcategories can be preset. For example, a first number of feature dimensions can be preset to be divided into 5 categories, and each category can be further divided into 6 subcategories. It is understood that the number of feature dimension categories and subcategories can be set according to actual circumstances and needs, and is not limited here.

[0058] According to some embodiments, such as Figure 3 As shown, a data processing method 300 is provided, including: step S301, obtaining a sample dataset for modeling, wherein each sample data in the sample dataset is high-dimensional sparse data and includes a first number of feature dimensions; step S302, dividing the first number of feature dimensions into multiple categories; step S303, selecting multiple feature dimensions from each of the multiple categories to obtain a second number of feature dimensions, including: step S304, for each of the multiple categories, dividing all feature dimensions included in that category into multiple subcategories; and step S305, selecting multiple feature dimensions from each of the multiple subcategories to constitute the second number of feature dimensions; and step S306, performing dimensionality reduction processing on each sample data in the sample dataset, including: step S307, deleting feature dimensions other than the second number of feature dimensions; and step S308, for each feature dimension in the second number of feature dimensions with a positive feature value, replacing the feature value of that feature dimension with a semantic vector corresponding to that feature dimension. The operations of steps S301-S308 are similar to those described above and will not be elaborated upon here.

[0059] In some embodiments, selecting multiple feature dimensions from each of the multiple subclasses includes: determining the feature coverage rate of each feature dimension included in the subclass in the sample dataset; and selecting the third number of feature dimensions with the highest feature coverage rate from the subclass as the multiple feature dimensions to be selected from the subclass. This ensures that the dimensionality-reduced data can cover the user's main features and avoids losing important features during the dimensionality reduction process.

[0060] Feature coverage refers to the ratio of the number of samples with positive values ​​for a specific feature dimension in a dataset to the total number of samples. For example, in a user behavior sample dataset, if 80% of users have positive values ​​for the feature dimension "whether they have installed application A", then the feature coverage for the feature dimension "whether they have installed application A" is 80%.

[0061] In some embodiments, multiple feature dimensions in each subclass can be sorted according to their feature coverage, and only the third largest number of feature dimensions are selected as the feature dimensions of the dimensionality-reduced sample data.

[0062] In some embodiments, the third quantity can be preset. For example, after sorting the feature dimensions in each subclass, only the top 10 feature dimensions are selected as the feature dimensions of the dimensionality-reduced sample data. It is understood that the third quantity can be set according to actual circumstances and needs, and is not limited here.

[0063] In some embodiments, the data processing method may further include: sorting the second number of feature dimensions according to categories. Thus, by sorting the feature dimensions according to categories, the order of the feature dimensions in the dimensionality-reduced sample data is unified, avoiding errors that could affect model performance when using the dimensionality-reduced sample dataset for model training.

[0064] In some embodiments, the ratio of the second quantity to the first quantity does not exceed 0.1%. For example, the feature dimension of a user behavior sample data can be 300,000 dimensions, while the feature dimension of the dimensionality-reduced sample data can be 300 dimensions or even lower. Thus, by performing dimensionality reduction processing on high-dimensional sparse sample data through the above operations and applying the dimensionality-reduced sample dataset for modeling, the problems of large modeling computation and low model training or inference efficiency can be solved.

[0065] In some embodiments, such as Figure 4 The unsupervised semantic representation model shown obtains the semantic vector for each feature dimension. For example, taking the feature dimension of application software as an example, the model can first be trained using an unlabeled feature dataset of application software, so that the model can represent the feature vector of application software as the corresponding semantic vector.

[0066] The trained model will generate the high-dimensional sparse feature vector x corresponding to the application software "App N". i (The dimension V-dim can be, for example, 300,000 dimensions) is input into the model through the input layer, and through the input layer (W) V×N The parameters of the input layer are used to obtain their corresponding semantic representations, and the corresponding semantic representation vector h is output through the output layer. i (The dimension N-dim can be, for example, 64 dimensions.)

[0067] In some embodiments, based on the above method, semantic vectors corresponding to feature dimensions with positive feature values ​​can be obtained from the dimensionality-reduced sample data, and these feature values ​​can be replaced using the corresponding semantic vectors. Therefore, using the sample dataset obtained based on the above method for model training can solve the problem of poor model generalization ability.

[0068] According to some embodiments, such as Figure 5 As shown, a model training method is provided, including: step S501, obtaining a sample dataset, wherein each sample data in the sample dataset is obtained based on the above data processing method and includes a corresponding sample label; step S502, initializing multiple parameters of the model, and for each sample data, performing the following operations: step S503, inputting the sample data into the model to obtain the model calculation result; and step S504, adjusting multiple parameters of the model based on the model calculation result and the corresponding sample label of the sample data.

[0069] Based on the embodiments of this disclosure, the dimensionality of the sample data is reduced by reconstructing the high-dimensional sparse sample data according to a preset category, thereby solving the problems of large computational load and low training efficiency during model training. At the same time, the feature dimensions with positive feature values ​​are replaced by corresponding semantic vectors, and the obtained sample dataset is used for model training, thereby solving the problem of poor model generalization ability.

[0070] According to some embodiments, such as Figure 6 As shown, a data processing apparatus 600 is provided, including: a first acquisition unit 610 configured to acquire a sample dataset for modeling, wherein each sample data in the sample dataset is high-dimensional sparse data and includes a first number of feature dimensions; a partitioning unit 620 configured to partition the first number of feature dimensions into multiple categories; a second acquisition unit 630 configured to select multiple feature dimensions from each of the multiple categories to acquire a second number of feature dimensions; and a dimensionality reduction processing unit 640 configured to perform dimensionality reduction processing on each sample data in the sample dataset, the dimensionality reduction processing unit 640 including: a deletion subunit 641 configured to delete feature dimensions other than the second number of feature dimensions; and a replacement subunit 642 configured to replace the feature value of each feature dimension with a semantic vector corresponding to the feature dimension for each feature dimension in the second number of feature dimensions with a positive feature value.

[0071] The operations of units 610-640 and subunits 641 and 642 in the data processing device 600 are similar to the operations of steps S201-S206 in the above data processing method, and will not be described in detail here.

[0072] According to some embodiments, such as Figure 7 As shown, a data processing apparatus 700 is also provided, wherein the second acquisition unit 730 includes: a division subunit 731 configured to divide all feature dimensions included in each of the multiple categories into multiple subcategories for each category of the multiple categories; and a selection subunit 732 configured to select multiple feature dimensions from each of the multiple subcategories for each category of the multiple categories to constitute a second number of feature dimensions.

[0073] The operations of units 710-740 and sub-units 731-732 and sub-units 741-742 in the data processing device 700 are similar to the operations of steps S301-S308 in the above data processing method, and will not be described in detail here.

[0074] According to some embodiments, the selection subunit may include: a determining module configured to determine the feature coverage of each feature dimension included in the subclass in the sample dataset; and a selection module configured to select a third number of feature dimensions with the highest feature coverage from the subclass, as a selection of multiple feature dimensions from the subclass.

[0075] According to some embodiments, the data processing apparatus may further include a sorting unit configured to sort the second number of feature dimensions according to categories.

[0076] According to some embodiments, the ratio of the second quantity to the first quantity does not exceed 0.1%.

[0077] According to some embodiments, such as Figure 8 As shown, a model training apparatus 800 is provided, comprising: a third acquisition unit 810 configured to acquire a sample dataset, wherein each sample data in the sample dataset is obtained based on the data processing method described above and includes a corresponding sample label; an initialization unit 820 configured to initialize multiple parameters of the model; and an execution unit 830 configured to perform the following sub-unit operations for each sample data, the execution unit 830 including: an input sub-unit 831 configured to input the sample data into the model to obtain model calculation results; and an adjustment sub-unit 832 configured to adjust multiple parameters of the model based on the model calculation results and the corresponding sample label of the sample data.

[0078] The operations of units 810-830 and sub-units 831-832 in the model training device 800 are similar to the operations of steps S501-S504 in the above model training method, and will not be described in detail here.

[0079] The collection, storage, use, processing, transmission, provision, and disclosure of user personal information involved in the technical solution disclosed herein comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0080] According to embodiments of this disclosure, an electronic device, a readable storage medium, and a computer program product are also provided.

[0081] refer to Figure 9 The present invention describes a structural block diagram of an electronic device 900 that can serve as a server or client of the present disclosure, which is an example of a hardware device that can be applied to various aspects of the present disclosure. The electronic device is intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0082] like Figure 9 As shown, device 900 includes a computing unit 901, which can perform various appropriate actions and processes based on a computer program stored in read-only memory (ROM) 902 or a computer program loaded from storage unit 908 into random access memory (RAM) 903. RAM 903 may also store various programs and data required for the operation of device 900. The computing unit 901, ROM 902, and RAM 903 are interconnected via bus 904. Input / output (I / O) interface 905 is also connected to bus 904.

[0083] Multiple components in device 900 are connected to I / O interface 905, including: input unit 906, output unit 907, storage unit 908, and communication unit 909. Input unit 906 can be any type of device capable of inputting information to device 900. Input unit 906 can receive input numerical or character information and generate key signal input related to user settings and / or function control of the electronic device, and can include, but is not limited to, a mouse, keyboard, touchscreen, trackpad, trackball, joystick, microphone, and / or remote control. Output unit 907 can be any type of device capable of presenting information, and can include, but is not limited to, a monitor, speaker, video / audio output terminal, vibrator, and / or printer. Storage unit 908 can include, but is not limited to, a hard disk and an optical disk. Communication unit 909 allows device 900 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks, and can include, but is not limited to, a modem, network card, infrared communication device, wireless communication transceiver, and / or chipset, such as Bluetooth. TM Devices, 802.11 devices, WiFi devices, WiMax devices, cellular communication devices and / or the like.

[0084] The computing unit 901 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 901 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 901 performs the various methods and processes described above, such as data processing methods. For example, in some embodiments, the data processing method may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 908. In some embodiments, part or all of the computer program may be loaded and / or installed on device 900 via ROM 902 and / or communication unit 909. When the computer program is loaded into RAM 903 and executed by the computing unit 901, one or more steps of the data processing method described above may be performed. Alternatively, in other embodiments, the computing unit 901 may be configured to perform data processing methods by any other suitable means (e.g., by means of firmware).

[0085] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0086] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0087] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0088] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0089] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with embodiments of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0090] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, servers in distributed systems, or servers incorporating blockchain technology.

[0091] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be performed in parallel, sequentially, or in a different order, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0092] While embodiments or examples of this disclosure have been described with reference to the accompanying drawings, it should be understood that the methods, systems, and devices described above are merely exemplary embodiments or examples, and the scope of the invention is not limited by these embodiments or examples, but only by the granted claims and their equivalents. Various elements in the embodiments or examples may be omitted or replaced by their equivalents. Furthermore, the steps may be performed in a different order than that described in this disclosure. Further, various elements in the embodiments or examples may be combined in various ways. Importantly, as the technology evolves, many elements described herein can be replaced by equivalents that appear after this disclosure.

Claims

1. A data processing method, comprising: Obtain a sample dataset for modeling, wherein each sample data in the sample dataset is high-dimensional sparse data and includes a first number of feature dimensions, the sample data is user behavior sample data, the sample data is a fixed-length structured vector, each position of the structured vector represents one of the first number of feature dimensions, and the feature dimension corresponding to each position of the structured vector corresponds to a specific mobile application and is used to characterize whether the user has the installation behavior characteristics of the mobile application. Based on the functions of mobile application software, the first number of feature dimensions are divided into multiple categories, and the multiple categories include at least two of system applications, entertainment and health, games and card games, daily shopping, and finance and current affairs. Multiple feature dimensions are selected from each of the multiple categories to obtain a second number of feature dimensions, wherein obtaining the second number of feature dimensions includes: For each of the multiple categories Divide all feature dimensions included in this category into multiple subclasses; and Selecting multiple feature dimensions from each of the multiple subclasses to constitute the second number of feature dimensions, wherein selecting multiple feature dimensions from each of the multiple subclasses includes: Determine the feature coverage of each feature dimension included in the subclass in the sample dataset; and Select the third number of feature dimensions with the highest feature coverage from this subclass, and use this as the selection of multiple feature dimensions from this subclass; and Dimensionality reduction is performed on each sample data in the sample dataset, and the dimensionality reduction process includes: Remove all feature dimensions other than the second number of feature dimensions from the sample data; and For each feature dimension in the second number of feature dimensions in the sample data with a positive feature value, the feature value of the feature dimension is replaced with a semantic vector corresponding to that feature dimension. Each sample data after the dimensionality reduction process is used for user behavior modeling. The semantic vector corresponding to the feature dimension is obtained by inputting the high-dimensional sparse feature vector of the specific mobile application software corresponding to the feature dimension into a pre-trained unsupervised semantic representation model for dimensionality reduction. The unsupervised semantic representation model is obtained by training on an unlabeled feature dataset, which includes high-dimensional sparse feature vectors corresponding to the mobile application software.

2. The method of claim 1, further comprising: The feature dimensions of the second quantity are sorted according to category.

3. The method as described in claim 1, wherein, The ratio of the second quantity to the first quantity does not exceed 0.1%.

4. A model training method in which, The model is used for user behavior modeling, and the method includes: Obtain a sample dataset, wherein each sample data in the sample dataset is obtained based on the method described in any one of claims 1 to 3 and includes a corresponding sample label; Initialize multiple parameters of the model, and For each sample data, perform the following operation: The sample data is input into the model to obtain the model calculation results; and Based on the model's calculation results and the corresponding sample labels of the sample data, several parameters of the model are adjusted.

5. A data processing apparatus, comprising: The first acquisition unit is configured to acquire a sample dataset for modeling, wherein each sample data in the sample dataset is high-dimensional sparse data and includes a first number of feature dimensions, the sample data is user behavior sample data, the sample data is a fixed-length structured vector, each position of the structured vector represents one of the first number of feature dimensions, and the feature dimension corresponding to each position of the structured vector corresponds to a specific mobile application and is used to characterize whether the user has the installation behavior characteristics of the mobile application. The segmentation unit is configured to divide the first number of feature dimensions into multiple categories based on the functions of the mobile application software. The multiple categories include at least two of system applications, entertainment and health, games and card games, daily shopping, and financial and political affairs. The second acquisition unit is configured to select multiple feature dimensions from each of the plurality of categories to obtain a second number of feature dimensions, wherein the second acquisition unit includes: The sub-unit is configured to, for each of the plurality of categories, divide all feature dimensions included in that category into multiple sub-categories; and The selection subunit is configured to, for each of the plurality of categories, select a plurality of feature dimensions from each of the plurality of subcategories to constitute the second number of feature dimensions, wherein the selection subunit includes: The determination module is configured to determine the feature coverage of each feature dimension included in the subclass in the sample dataset; and The selection module is configured to select the third number of feature dimensions with the highest feature coverage from that subclass, as a selection of multiple feature dimensions from that subclass; and A dimensionality reduction processing unit is configured to perform dimensionality reduction processing on each sample data of the sample dataset, the dimensionality reduction processing unit comprising: The deletion subunit is configured to delete feature dimensions other than the second number of feature dimensions in the sample data; and The replacement subunit is configured to replace the feature value of each feature dimension with a corresponding semantic vector for each feature dimension in the second number of feature dimensions of the sample data that has a positive feature value. Each sample data after the dimensionality reduction is used for user behavior modeling. The semantic vector corresponding to the feature dimension is obtained by inputting the high-dimensional sparse feature vector of the specific mobile application software corresponding to the feature dimension into a pre-trained unsupervised semantic representation model for dimensionality reduction. The unsupervised semantic representation model is obtained by training on an unlabeled feature dataset, which includes the high-dimensional sparse feature vector corresponding to the mobile application software.

6. The apparatus of claim 5, further comprising: The sorting unit is configured to sort the second number of feature dimensions according to categories.

7. The apparatus of claim 5, wherein, The ratio of the second quantity to the first quantity does not exceed 0.1%.

8. A model training apparatus, wherein, The model is used for user behavior modeling, and the device includes: The third acquisition unit is configured to acquire a sample dataset, wherein each sample data in the sample dataset is obtained based on the method described in any one of claims 1 to 3 and includes a corresponding sample label; An initialization unit is configured to initialize multiple parameters of the model, and An execution unit is configured to perform the operations of the following subunits for each sample data, the execution unit comprising: The input subunit is configured to input the sample data into the model to obtain model calculation results; and The adjustment subunit is configured to adjust multiple parameters of the model based on the model calculation results and the corresponding sample labels of the sample data.

9. An electronic device, comprising: At least one processor; as well as A memory that is communicatively connected to the at least one processor; in The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-4.

10. A non-transitory computer readable storage medium having stored thereon computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-4.

11. A computer program product comprising a computer program, wherein, The computer program, when executed by a processor, implements the method of any one of claims 1-4.

Images