CN115114515A

CN115114515A - Content recommendation method based on user interest and terminal equipment

Info

Publication number: CN115114515A
Application number: CN202110307500.XA
Authority: CN
Inventors: 邢超; 赵洋; 赵路德; 陈少杰
Original assignee: Huawei Technologies Co Ltd
Current assignee: Huawei Technologies Co Ltd
Priority date: 2021-03-23
Filing date: 2021-03-23
Publication date: 2022-09-27
Also published as: WO2022199494A1

Abstract

The application discloses a content recommendation method and terminal equipment based on user interests, which are used for improving the accuracy and efficiency of content recommendation based on the user interests on the basis of protecting user privacy data. Acquiring a plurality of user operation behavior data input by a user when the user uses a target application program for one time or a plurality of times in a set time length through terminal equipment; carrying out desensitization processing on the collected multiple user operation behavior data; and sending the desensitized plurality of user operation behavior data to a server so that the server analyzes the desensitized plurality of user operation behavior data to obtain a subject interest list of the user using the target application program. And the terminal equipment receives a theme interest list sent by the server, and displays a first recommendation interface when the user starts the target application program, wherein the first recommendation interface comprises at least one item of recommendation content determined according to the theme interest list.

Description

Content recommendation method based on user interest and terminal equipment

Technical Field

The present application relates to the field of data processing technologies, and in particular, to a content recommendation method and a terminal device based on user interests.

Background

With the development of internet technology, more and more applications or systems focus more on thousands of content recommendation technologies in order to improve user experience. The content recommendation system based on the user interest is often required to be established on the basis of massive user operation behavior data, and accordingly the problem of protecting the privacy data of the user is brought. The privacy data of the user includes, for example, personal information of the user, personalized interests of the user, and the like.

In the prior art, there is a technical solution for protecting private data of a user by using a secure encryption technology (e.g. a homomorphic encryption technology). Although the secure encryption technology is adopted for the private data of the user when the terminal device and the server side transmit and store the user operation behavior data, the privacy leakage security risk still exists in the user operation behavior data transmission process. In the prior art, a recommendation model is trained by using a distributed federal learning technology, which can protect user privacy by keeping user operation behavior data at a terminal device side, but the terminal device side is required to have the capability of training the model, so that the performance requirement on the terminal device side is high, and the recommendation efficiency is low.

Therefore, there is a great challenge in improving the accuracy and efficiency of content recommendation based on user interests on the basis of protecting user privacy data.

Disclosure of Invention

The embodiment of the application provides a content recommendation method and terminal equipment based on user interests, which are used for improving the accuracy and efficiency of content recommendation based on the user interests on the basis of protecting user privacy data.

In a first aspect, an embodiment of the present application provides a content recommendation method based on user interests, which may be applied to a terminal device. The method comprises the following steps: collecting a plurality of user operation behavior data input by a user when the user uses a target application program for one time or a plurality of times in a set time length; desensitizing the collected user operation behavior data, wherein the desensitizing treatment is to filter out privacy data related to the user in the user operation behavior data; and sending the desensitized plurality of user operation behavior data to a server so that the server analyzes the desensitized plurality of user operation behavior data to obtain a subject interest list of the user using the target application program.

According to the method provided by the embodiment of the application, the terminal equipment can collect the user operation behavior data of the user in the target application program, but the user privacy data can be uploaded to the server after desensitization processing is carried out on the terminal equipment, so that the server cannot obtain the user privacy data, the user privacy can be well protected, and the user experience is improved. The target application program may be any application program, such as a browser, suitable for content recommendation based on user interests provided in the embodiment of the present application.

In one possible design, the collecting, by the terminal device, a plurality of user operation behavior data input by the user when the user uses the target application one or more times in a set time period may be implemented as: when an instruction of the user for starting the target application program is detected, starting the target application program; after the target application program is started, acquiring at least one operation data executed by the user on the target application program; closing the target application program when detecting the instruction of the user for exiting the target application program; and storing at least one operation data collected from the starting to closing process of the target application program as a group of user operation behavior data.

In the design, the terminal device collects the user operation behavior data when the user uses the target application program, desensitizes the user operation behavior data, and uploads the desensitized user operation behavior data to the server. Therefore, the server can conveniently perform big data analysis based on the desensitization-processed user operation behavior data to obtain the user interest of the user in the target application program.

In one possible design, the terminal device performs random replacement of the user operation behavior data under the same interest topic based on a differential privacy algorithm for one or more user operation behavior data in the plurality of user operation behavior data, wherein the interest topic is determined according to the topic interest table; and stripping the user information contained in the plurality of user operation behavior data.

In the design, the terminal device randomly replaces the user operation behavior data based on the differential privacy algorithm, so that the effect of covering the real user operation behavior data can be realized, and the purpose of protecting the user privacy can be further achieved. In addition, before the terminal equipment uploads the user operation behavior data to the server, the purpose that the server cannot collect the user privacy data can be achieved by stripping the user information, and therefore the privacy and the safety of the user data can be guaranteed.

In a possible design, before the random replacement of the user operation behavior data under the same interest topic is performed based on the differential privacy algorithm, the terminal device may further determine a sequence length of each user operation behavior data; and performing truncation and compensation processing on the user operation behavior data according to a preset value to obtain the user operation behavior data with the appointed sequence length.

In the design, if the operation data included in the user operation behavior data is too little, that is, the sequence length is short, more accurate user interest cannot be analyzed from the user operation behavior data. The problem of excessive calculation amount is caused by excessive operation data contained in the user operation behavior data, that is, the sequence length is long. Therefore, the user operation behavior data with uniform sequence length is obtained by sampling the user operation behavior data according to the preset value, so that the processing efficiency of desensitization processing on the user operation behavior data can be improved, and the efficiency and accuracy of statistical analysis on the user operation behavior data by the server are improved.

Optionally, the terminal device performs truncation and compensation processing on the user operation behavior data according to a preset value to obtain the user operation behavior data with a specified sequence length, which may be specifically implemented as follows: if the sequence length of the user operation behavior data is smaller than the preset value, supplementing the user operation behavior data with the predefined user operation behavior data with the target length to obtain the user operation behavior data with the specified sequence length; or if the sequence length of the user operation behavior data is larger than the preset value, cutting off the target length of the user operation behavior data to obtain the user operation behavior data with the specified sequence length; and the target length is the absolute value of the difference value between the sequence length of the user operation behavior data and a preset value.

In the design, a specific scene of sampling according to a preset value is given, the sequence length of the user operation behavior data is judged, if the sequence length is smaller than the preset value of the user operation behavior data, predefined default user operation behavior data can be adopted for complementing, and the user operation behavior data with the sequence length larger than the preset value can be randomly truncated. Therefore, after the user operation behavior data are sampled according to the preset value, the user operation behavior data with uniform sequence length can be obtained, so that desensitization treatment can be conveniently carried out.

In a second aspect, an embodiment of the present application provides a content recommendation method based on user interests, which may be applied in a server. The method comprises the following steps: receiving a plurality of user operation behavior data after desensitization processing sent by one or more terminal devices; the desensitized user operation behavior data is obtained by acquiring a plurality of user operation behavior data input by a user when the user uses a target application program for one time or a plurality of times in a set time length for the one or a plurality of terminal devices and desensitizing the acquired plurality of user operation behavior data; the desensitization processing is to filter out privacy data related to the user in the user operation behavior data; analyzing the desensitized multiple user operation behavior data to obtain a subject interest list of the user using the target application program; and sending the topic interest list to the one or more terminal devices.

In the method, the server has better computing capacity compared with the terminal equipment, and the server can comprehensively analyze the user operation behavior data sent by the plurality of terminal equipment to perform group-based analysis on the user operation behavior data, so that a hot interest topic which meets timeliness can be obtained, and the method can also be understood as an interest topic which is currently compared and is concerned by most users and generates a topic interest list. And the server can send the topic interest list to the terminal equipment for the terminal equipment to perform real-time recommendation by combining the topic interest list so as to improve the user experience.

In a possible design, the server analyzes the desensitized multiple user operation behavior data to obtain a topic interest list of the user using the target application program, which may be specifically implemented as: inputting the desensitized plurality of user operation behavior data into a pre-constructed theme interest model to perform unsupervised learning on the desensitized plurality of user operation behavior data; and obtaining a topic interest table output by the pre-constructed topic interest model.

In the design, a topic interest table can be obtained according to a large amount of user operation behavior data sent by a plurality of terminal devices through a pre-constructed topic interest model, and the obtained topic interest table can better reflect the interest topics which are currently concerned by most users.

In a third aspect, an embodiment of the present application provides a content recommendation method based on user interests, which may be applied to a terminal device. The method comprises the following steps: receiving a theme interest table sent by a server, wherein the theme interest table is obtained by analyzing the desensitized operation behavior data of a plurality of users by the server; the desensitized user operation behavior data is obtained by acquiring a plurality of user operation behavior data input by a user when the user uses a target application program for one time or a plurality of times in a set time length by one or a plurality of terminal devices and desensitizing the acquired plurality of user operation behavior data; the desensitization processing is to filter out privacy data related to the user in the user operation behavior data; when an instruction of the user for starting the target application program is detected, starting the target application program and displaying a first recommendation interface, wherein the first recommendation interface comprises at least one item of recommendation content; the at least one item of recommended content is determined from the topic interest table.

In the method, the terminal equipment can perform real-time recommendation by combining the topic interest list sent by the server. In this scenario, if the user belongs to a new user or the terminal device does not store the historical user interests of the user, the terminal device may recommend according to the topic interest table, and by comparing the recommendation of the content corresponding to the interest topics concerned by most users at present, it is possible to avoid cold start recommendation in the target application program, that is, the target application program recommends some content corresponding to cold interest topics, and even cannot recommend the content, so that the browsing interest of the user cannot be aroused.

In a possible design, the launching the target application and displaying the first recommendation interface may be implemented as: starting the target application program; displaying a first recommendation interface after the target application program is started; taking one or more interest topics contained in the topic interest table as user interests, acquiring at least one item of recommended content according to the user interests, and displaying the acquired at least one item of recommended content in the first recommendation interface; each interest topic has an associated weight value, and the larger the weight value associated with an interest topic is, the higher the proportion of related content containing the interest topic in the recommended content is.

In the design, content recommendation is performed according to the weight values associated with the interest topics contained in the topic interest table, so that the higher proportion of the related recommended content corresponding to the more popular interest topics in the recommendation interface can be realized, the possibility of interest of the user can be improved, and the user experience can be improved.

In one possible design, after the target application is started and the first recommendation interface is displayed, the method is implemented as follows: receiving and collecting one or more user operation behavior data input by a user when the user uses the target application program; when an instruction of refreshing the first recommendation interface by the user is detected, displaying a second recommendation interface; and the recommended content contained in the second recommendation interface is determined according to the one or more user operation behavior data and the topic interest table.

In the design, when the application is implemented, after the target application program obtains the real-time operation data of the user, the interest theme concerned by the user can be obtained through analysis of the real-time operation data of the user, so that the interest of the user can be adjusted in time, and a recommendation interface matched with the interest of the user can be displayed in time.

In one possible design, the displaying the second recommendation interface may be implemented as: determining one or more corresponding interest topics according to the one or more user operation behavior data, and distributing associated weight values for the interest topics; taking one or more interest topics corresponding to the user operation behavior data and one or more interest topics contained in the topic interest table as user interests, acquiring at least one item of recommended content according to the user interests, and displaying the acquired at least one item of recommended content in the second recommendation interface; each interest topic included in the topic interest table has an associated weight value, and the larger the weight value associated with an interest topic is, the higher the proportion of related content of the interest topic included in the recommended content is.

In the design, by combining the real-time user operation behavior data of the user and the topic interest table generated by the server, the personal interests of the user can be considered on the basis of the current popular interest topics, so that recommended content more conforming to the interests of the user can be obtained, and the use experience of the user can be improved.

In one possible design, the obtaining at least one item of recommended content according to the user interest may be implemented as: searching for recommended content corresponding to the user interest from local cache content; and/or acquiring recommended content corresponding to the user interest from a content providing server providing the recommended content corresponding to the user interest.

In the design, after the terminal device determines the user interest, recommended content related to the user interest, such as hot articles, hot news and the like, can be acquired in multiple possible ways to improve diversity of the recommended content.

In a fourth aspect, an embodiment of the present application further provides a terminal device, including: one or more processors; one or more memories; the one or more memories for storing one or more computer programs and data information; wherein the one or more computer programs comprise instructions; the instructions, when executed by the one or more processors, cause the terminal device to perform the method of any one of the first aspects above, or to perform the method of any one of the third aspects above.

In a fifth aspect, an embodiment of the present application further provides a server, including: one or more processors; one or more memories; the one or more memories for storing one or more computer programs and data information; wherein the one or more computer programs comprise instructions; the instructions, when executed by the one or more processors, cause the server to perform the method of any of the second aspects above.

In a sixth aspect, an embodiment of the present application further provides a communication system, including: a terminal device and a server; the terminal device may perform the steps of the terminal device in the method as provided in the first aspect above, or perform the steps of the terminal device in the method as provided in the third aspect above; the server may perform the steps of the server in the method as provided in the second aspect above.

In a seventh aspect, the present application provides a computer-readable storage medium, which stores a computer program (which may also be referred to as code or instructions) and when the computer program runs on a computer, causes the computer to perform the method in any one of the above-mentioned possible implementations of the first aspect, or perform the method in any one of the above-mentioned possible implementations of the second aspect, or perform the method in any one of the above-mentioned possible implementations of the third aspect.

In an eighth aspect, an embodiment of the present application provides a computer program product, where the computer program product includes: a computer program (also referred to as code, or instructions), which when executed, causes a computer to perform the method of any of the possible implementations of the first aspect described above, or the method of any of the possible implementations of the second aspect described above, or the method of any of the possible implementations of the third aspect described above.

In a ninth aspect, an embodiment of the present application further provides a graphical user interface on a terminal device, where the terminal device has a display screen, one or more memories, and one or more processors, where the one or more processors are configured to execute one or more computer programs stored in the one or more memories, and the graphical user interface includes a graphical user interface displayed when the terminal device executes any of the possible implementation manners of the first aspect of the embodiment of the present application, or a graphical user interface displayed when the terminal device executes any of the possible implementation manners of the third aspect of the embodiment of the present application.

For the advantages of any one of the fourth aspect to the ninth aspect, please refer to various possible designs of the first aspect to the third aspect, which are not repeated herein.

Drawings

Fig. 1 is an application scene diagram of a content recommendation method based on user interests according to an embodiment of the present application;

fig. 2a is a schematic diagram of a hardware architecture of a terminal device according to an embodiment of the present application;

fig. 2b is a block diagram of a software structure of a terminal device according to an embodiment of the present disclosure;

fig. 3 is a schematic structural diagram of a content recommendation method based on user interests according to an embodiment of the present application;

FIG. 4 is a schematic view of a user interface of a content recommendation method based on user interests according to an embodiment of the present application;

FIG. 5 is a schematic diagram of a radar map of a user interest topic provided by an embodiment of the present application;

fig. 6 is a flowchart illustrating a content recommendation method based on user interests according to an embodiment of the present application;

fig. 7 is a second schematic diagram of a user interface of a content recommendation method based on user interests according to an embodiment of the present application;

fig. 8a is a third schematic view of a user interface of a content recommendation method based on user interests according to an embodiment of the present application;

FIG. 8b is a fourth schematic view of a user interface of a content recommendation method based on user interests according to an embodiment of the present application;

fig. 8c is a fifth schematic view of a user interface of a content recommendation method based on user interests according to an embodiment of the present application;

fig. 9 is a schematic structural diagram of a terminal device or a server according to an embodiment of the present application.

Detailed Description

With the rapid development of society, terminal devices such as mobile phones are becoming more and more popular. The terminal equipment not only has a communication function, but also has strong processing capability, storage capability, a photographing function and the like. The terminal device executes a corresponding application program through an operating system (e.g., an android operating system), and a user can make a call, send a short message, browse a webpage, take a picture, play a game, watch a video, and the like by using the terminal device. In some Applications (APPs) with data in different fields of interest, the terminal device may recommend content according to user interest, for example, when a user performs user operation behaviors such as searching in a browser and browsing a web page, when a small video APP browses a small video, or when a shopping APP makes a shopping, the terminal device may recommend content that the user may be interested in according to user operation behaviors such as a search word, a search history record, and a currently browsed content of the user. The user interest may be a customized topic interest category in the application program, or a commonly used topic interest category, for example, the topic interest category in the browser may be a sports category, a financial category, a real-time category, and the topic interest category in the shopping APP may be a clothing category, a living goods category, a food category, and the like.

In combination with the description in the background art, the content recommendation system based on the user interest needs to collect a large amount of user operation behavior data to complete content recommendation, however, at present, on the basis of protecting the user privacy, there is no good solution for improving the recommendation accuracy and recommendation efficiency of the content recommendation system based on the user interest.

In view of this, the present application provides a content recommendation method based on user interests, which collects user operation behavior data at a terminal device side, and uploads the user operation behavior data obtained after stripping user sensitive privacy data to a server side. And the server side constructs a theme interest list according to a large amount of user operation behavior data uploaded by the plurality of terminal equipment sides and returns the theme interest list to the terminal equipment. The terminal device may determine the user interest by combining the topic interest table issued by the server side, the real-time operation behavior data of the user on the terminal device, the historical user interest and other factors. Finally, the terminal device may request the server side for the relevant content of the user interest according to the determined user interest, or may obtain the relevant content of the user interest from a local cache of the terminal device, so as to implement real-time content recommendation on the terminal device.

The embodiments of the present application will be described in detail below with reference to the accompanying drawings.

Fig. 1 is an application scenario diagram of a content recommendation method based on user interests according to an embodiment of the present application. The application scenario may include the terminal device 110, the server 120, and the database 130, the terminal device 110 may have an application installed therein, and the server 120 may be a background server communicating with the terminal device or a separate server for mining potential objects. The application may be a web application or an application pre-installed in the terminal device 110, and the application in the present application may be any type of application that can recommend content, such as a browser application, a small video application, and a shopping application. Both terminal device 110 and server 120 may access database 130 and store access logs generated during user access in database 130. The database 130 may be disposed on the server 120, or may be disposed independent of the server 120, for example, the database 130 may be implemented by a server cluster, a cloud server, or a distributed storage server. It should be noted that the number and types of the terminal devices 110, the server 120, and the database 130 included in the application scenario are not limited in the present application, and for example, there may be a plurality of terminal devices 110 shown in fig. 1.

For example, when the current user starts an application program in the terminal device 110 and performs a user operation on the application program to generate user operation behavior data, the user operation behavior data may be, for example, operation data of a user performing search, browsing, and the like in a browser, and the terminal device 110 may perform content recommendation according to the user operation behavior data and a topic interest list issued by the server 120. After the terminal device 110 closes the application program, desensitizes the user operation behavior data, and uploads the desensitized user operation behavior data to the server 120, and then the server 120 may generate or update the topic interest list according to a large amount of user operation behavior data, and then returns the topic interest list to the terminal device 110. In addition, the server 120 may also store the generated topic interest table or the received user operation behavior data in the database 130.

It is understood that the terminal device 110 of the embodiment of the present application may be a mobile device, a tablet computer, a wearable device (e.g., a watch, a bracelet, a helmet, a headset, etc.), an in-vehicle device, an Augmented Reality (AR)/Virtual Reality (VR) device, a notebook computer, an ultra-mobile personal computer (UMPC), a netbook, a Personal Digital Assistant (PDA), a smart home device (e.g., a smart television, a smart speaker, a smart camera, etc.), and the like. It is to be understood that the specific type of the terminal device 110 is not limited in any way in the embodiments of the present application.

Terminal device 110 to which embodiments of the present application may be applied, exemplary embodiments include, but are not limited to, piggy-backing

Or other operating system. The portable terminal device described above may also be other portable terminal devices such as a Laptop computer (Laptop) with a touch sensitive surface, e.g. a touch panel, etc.

Fig. 2a shows a schematic diagram of a possible hardware structure of the terminal device. Wherein the terminal device 110 includes: radio Frequency (RF) circuitry 210, a power supply 220, a processor 230, a memory 240, an input unit 250, a display unit 260, an audio circuit 270, a communication interface 280, and a wireless fidelity (Wi Fi) module 290. Those skilled in the art will appreciate that the hardware structure of the terminal device shown in fig. 2a does not constitute a limitation of the terminal device, and the terminal device provided in the embodiments of the present application may include more or less components than those shown, may combine two or more components, or may have a different configuration of components. The various components shown in fig. 2a may be implemented in hardware, software, or a combination of hardware and software, including one or more signal processing and/or application specific integrated circuits.

The following describes each component of the terminal device 110 in detail with reference to fig. 2 a:

the RF circuit 210 may be used for receiving and transmitting data during a communication or conversation. Specifically, the RF circuit 210 sends downlink data of the base station to the processor 230 for processing; and in addition, sending the uplink data to be sent to the base station. Generally, the RF circuit 210 includes, but is not limited to, an antenna, at least one amplifier, a transceiver, a coupler, a Low Noise Amplifier (LNA), a duplexer, and the like.

In addition, the RF circuitry 210 may also communicate with networks and other devices via wireless communications. The wireless communication may use any communication standard or protocol, including but not limited to global system for mobile communications (GSM), General Packet Radio Service (GPRS), Code Division Multiple Access (CDMA), Wideband Code Division Multiple Access (WCDMA), Long Term Evolution (LTE), email, Short Message Service (SMS), etc.

Wi Fi technology belongs to short distance wireless transmission technology, terminal equipment 110 can connect Access Point (AP) through Wi Fi module 290 to realize the visit of data network. The Wi-Fi module 290 may be used for receiving and transmitting data during communication.

The terminal device 110 may be physically connected to other devices via the communication interface 280. Optionally, the communication interface 280 is connected to the communication interface of the other device through a cable, so as to implement data transmission between the terminal device 110 and the other device.

In the embodiment of the present application, the terminal device 110 can implement a communication service, and implement interaction with a server side, so that the terminal device 110 needs to have a data transmission function, that is, the terminal device 110 needs to include a communication module inside. Although fig. 2a shows communication modules such as the RF circuit 210, the Wi-Fi module 290, and the communication interface 280, it is understood that at least one of the above components or other communication modules (such as a bluetooth module) for realizing communication exists in the terminal device 110 for data transmission.

For example, when the terminal device 110 is a mobile phone, the terminal device 110 may include the RF circuit 210 and may further include the Wi-Fi module 290; when the terminal device 110 is a computer, the terminal device 110 may include the communication interface 280 and may further include the Wi-Fi module 290; when the terminal device 110 is a tablet computer, the terminal device 110 may include the Wi-Fi module.

The memory 240 may be used to store software programs and modules. The processor 230 executes various functional applications and data processing of the terminal device 110 by executing software programs and modules stored in the memory 240. Alternatively, the memory 240 may mainly include a program storage area and a data storage area. The storage program area may store an operating system (mainly including a kernel layer, a system layer, an application framework layer, an application layer, and other software programs or modules corresponding to each other). The application program layer can contain various applications, and content recommendation based on user interests can be realized by adopting the method provided by the embodiment of the application in the applications capable of performing recommendation.

Further, the memory 240 may include high speed random access memory and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other volatile solid state storage device.

The input unit 250 may be used to receive editing operations of a plurality of different types of data objects such as numeric or character information input by a user and to generate key signal inputs related to user settings and function control of the terminal device 110. Alternatively, the input unit 250 may include a touch panel 251 and other input devices 252.

The touch panel 251, also referred to as a touch screen, may collect a touch operation performed by a user on or near the touch panel 251 (for example, an operation performed by the user on or near the touch panel 251 using any suitable object or accessory such as a finger, a stylus, etc.), and drive a corresponding connection device according to a preset program.

Optionally, the other input devices 252 may include, but are not limited to, one or more of a physical keyboard, function keys (e.g., volume control keys, switch keys, etc.), a trackball, a mouse, a joystick, and the like.

The display unit 260 may be used to display information input by a user or information provided to a user and various menus of the terminal device 110. The display unit 260 is a display system of the terminal device 110, and is configured to present an interface to implement human-computer interaction. The display unit 260 may include a display panel 261. Alternatively, the display panel 261 may be configured in the form of a Liquid Crystal Display (LCD), an organic light-emitting diode (OLED), or the like. In this embodiment of the application, for example, a visualization page corresponding to an operation of a user on a terminal device may be displayed by the display unit 260, for example, after the user inputs a search term, the display unit 260 displays an information stream, a web page, and the like corresponding to the search term.

The processor 230 is a control center of the terminal device 110, connects various components using various interfaces and lines, and implements various functions of the terminal device 110 and processes data by running or executing software programs and/or modules stored in the memory 240 and calling data stored in the memory 240, thereby implementing various services based on the terminal device. In the embodiment of the present application, the processor 230 is configured to implement the method provided in the embodiment of the present application, so as to recommend content of interest to a user more accurately.

The terminal device 110 also includes a power supply 220 (such as a battery) for powering the various components. Optionally, the power supply 220 may be logically connected to the processor 230 through a power management system, so as to implement functions of managing charging, discharging, power consumption, and the like through the power management system.

As shown in fig. 2a, the terminal device 110 further comprises an audio circuit 270, a microphone 271 and a speaker 272, which may provide an audio interface between a user and the terminal device 110. The audio circuit 270 may be configured to convert audio data into a signal that can be recognized by the speaker 272 and to transmit the signal to the speaker 272 for conversion by the speaker 272 into an audio signal for output. The microphone 271 is used for collecting external sound signals (such as voice of a human being, other sounds, etc.), converting the collected external sound signals into signals that can be recognized by the audio circuit 270, and sending the signals to the audio circuit 270. The audio circuit 270 may also be used to convert signals sent by the microphone 271 into audio data, and output the audio data to the RF circuit 220 for transmission to, for example, another terminal, or output the audio data to the memory 240 for further processing.

Although not shown, the terminal device 110 may further include at least one sensor, a camera, and the like, which are not described in detail herein.

The Operating System (OS) according to the embodiment of the present application is the most basic system software running on the terminal device 110. Taking a smartphone as an example, the operating system may be an android (android) system or an IOS system. The following embodiments are described taking the android system as an example. Those skilled in the art will appreciate that other operating systems may be implemented in a similar manner.

The software system of the terminal device 110 may adopt a layered architecture, an event-driven architecture, a micro-core architecture, a micro-service architecture, or a cloud architecture. In the embodiment of the present application, a software structure of the terminal device 110 is exemplarily described by taking an android system adopting a hierarchical architecture as an example.

Fig. 2b shows a software structure block diagram of the android system provided in the embodiment of the present application. The layered architecture divides the software into several layers, each layer having a clear role and division of labor. The layers communicate with each other through a software interface. In some embodiments, the android system is divided into five layers, an application layer, an application framework (framework) layer, an android runtime (android runtime) and system library, a hardware abstraction layer, and a kernel layer from top to bottom.

The application layer is the top layer of the operating system and may include a series of application packages. As shown in fig. 2b, the application layer may include a native application of the operating system and a third-party application, where the native application of the operating system may include a User Interface (UI), a browser, a camera, settings, a cell phone manager, music, a short message, a call, and the like, and the third-party application may include a map, a shopping APP, a small video APP, and the like. The application mentioned below may be a native application of an operating system installed when the terminal device 110 leaves a factory, or may be a third-party application downloaded from a network or acquired from another terminal device 110 by a user during use of the terminal device 110.

In some embodiments of the present application, the application layer may be used to implement presentation of an editing interface, which may be used for user operations. For example, the user may perform a user operation behavior such as inputting a search word on an editing interface correspondingly presented by the browser.

In a possible implementation manner, the application program may be developed using java language, and completed by calling an Application Programming Interface (API) provided by an application framework layer, and a developer may interact with a bottom layer (e.g., a hardware abstraction layer, a kernel layer, etc.) of an operating system through the application framework layer to develop its own application program. The application framework layer is primarily a series of services and management systems for the operating system.

The application framework layer provides an application programming interface and a programming framework for the application of the application layer. The application framework layer includes some predefined functions. As shown in FIG. 2b, the application framework layer may include a window manager, a content provider, a view system, a phone manager, a resource manager, a notification manager, and the like.

The window manager is used for managing window programs. The window manager can obtain the size of the display screen, judge whether a status bar exists, lock the screen, intercept the screen and the like. The content provider is used to store and retrieve data and make it accessible to applications. The data may include video, images, audio, calls made and received, browsing history and bookmarks, phone books, etc. The view system includes visual controls such as text controls that display text, picture controls that display pictures, and the like. The view system may be used to build applications. The display interface may be composed of one or more views. The telephone manager is used to provide communication functions of the terminal device 110, such as management of call state display (including connection, hang-up, etc.). The resource manager provides various resources, such as localized strings, icons, pictures, layout files, video files, etc., to the application.

In some embodiments of the present application, the application framework layer is mainly responsible for invoking a service interface for communicating with the hardware abstraction layer, so as to transfer an operation request operated by a user to the hardware abstraction layer, where the operation request may include an operation request corresponding to a certain APP opened by the user, or may include an operation request corresponding to a search term entered by the user at the certain APP, and the like. And the hardware abstraction layer generates corresponding content recommendation service according to the operation request transmitted by the application program layer.

Illustratively, the content recommendation service may include a data acquisition module, a data calibration module, a real-time recommendation module, a privacy protection module, and the like for implementing the methods provided herein. The data acquisition module is used for acquiring the user operation behavior of the user on the client on the terminal device to obtain the user operation behavior data. The data calibration module is used for preprocessing the user operation behavior data acquired by the data acquisition module to obtain the user operation behavior data with uniform sequence length. The privacy protection module is used for desensitizing the collected user operation behavior data, stripping or replacing the user operation behavior data related to the user privacy data in the user operation behavior data, and the like, so that the user operation behavior data which do not show the user privacy are obtained, the desensitized user operation behavior data are transmitted to the server side, and the desensitized user operation behavior data are used for constructing a subject interest model and generating a subject interest list. And the real-time recommendation module is used for carrying out real-time content recommendation according to the determined user interest.

An android runtime (android runtime) includes a core library and a virtual machine. The android runtime is responsible for scheduling and management of the android system. The core library of the android system comprises two parts: one part is a function which needs to be called by java language, and the other part is a core library of android.

The application layer and the application framework layer run in a virtual machine. Taking java as an example, the virtual machine executes java files of the application layer and the application framework layer as binary files. The virtual machine is used for performing the functions of object life cycle management, stack management, thread management, safety and exception management, garbage collection and the like.

The system library may include a plurality of functional modules. For example: surface managers (surface managers), media libraries (media libraries), three-dimensional graphics processing libraries (e.g., OpenGL ES), two-dimensional (2D) graphics engines (e.g., SGL), and the like.

The surface manager is used to manage the display subsystem and provide fusion of the 2D and 3D layers for multiple applications. The media library supports a variety of commonly used audio, video format playback and recording, and still image files, among others. The media library may support a variety of audio-video encoding formats, such as: MPEG4, H.264, MP3, AAC, AMR, JPG, PNG, etc. The three-dimensional graphic processing library is used for realizing three-dimensional graphic drawing, image rendering, synthesis, layer processing and the like. The 2D graphics engine is a drawing engine for 2D drawing.

A Hardware Abstraction Layer (HAL) is a support for an application framework layer, and is an important link for connecting the application framework layer and a kernel layer, and can provide services for developers through the application framework layer.

Illustratively, the functions of the content recommendation service in the embodiments of the present application may be implemented by configuring a first process at a hardware abstraction layer, and the first process may be a sub-process separately built in the hardware abstraction layer. The first process may include modules such as a content recommendation service configuration interface, a content recommendation service controller, and the like. Wherein, the content recommendation service configuration interface is a service interface for communicating with the application framework layer. The content recommendation service controller is configured to monitor a configuration interface of the content recommendation service, for example, whether the content recommendation service needs to be authenticated or not is controlled, and is also responsible for monitoring whether data input in the terminal device 110 needs to be cached or updated, and when the input data needs to be cached or updated, the application framework layer may be notified to cache or update corresponding data, so as to ensure that the latest data is displayed on the display interface. The hardware abstraction layer may further include a daemon process, where the daemon process may be used to cache data in the first process, and the daemon process may also be a sub-process separately constructed in the hardware abstraction layer.

The kernel layer may be a Linux kernel (Linux kernel) layer, which is an abstraction layer between hardware and software. The kernel layer has a plurality of drivers associated with the terminal device 110, including at least a display driver; linux-based frame buffer drivers; a keyboard drive and a mouse drive as input devices; flash drive based on memory technology equipment; audio driving; bluetooth drive, etc., and the embodiments of the present application do not set any limit to this. The Linux kernel layer is used for providing core system services of the operating system, and the security, the memory management, the process management, the network protocol stack, the driving model and the like are all realized based on the Linux kernel.

Typically, terminal device 110 may run multiple applications simultaneously. It is simpler, and an application can correspond to a process, and more complicated, an application can correspond to a plurality of processes. Each process is provided with a process number (process ID).

With reference to the description of the hardware structure of the terminal device 110 in fig. 2a and the description of the software framework of the terminal device 110 in fig. 2b, the following description will exemplarily illustrate the operation principle of the software and the hardware of the terminal device 110 for executing the content recommendation method based on user interests in the embodiment of the present application with respect to the context of content recommendation based on user interests.

It should be understood that "at least one" in the embodiments of the present application means one or more, "a plurality" means two or more. "and/or" describes the association relationship of the associated objects, meaning that there may be three relationships, e.g., a and/or B, which may mean: a alone, both A and B, and B alone, where A, B may be singular or plural. The character "/" generally indicates that the former and latter associated objects are in an "or" relationship. "at least one of the following" or similar expressions refer to any combination of these items, including any combination of the singular or plural items. For example, at least one (one) of a, b, or c, may represent: a, b, c, a and b, a and c, b and c, or a, b and c, wherein a, b and c can be single or multiple.

The embodiments of the present application relate to a plurality of numbers greater than or equal to two.

In addition, it is to be understood that the terms first, second, etc. in the description of the present application are used for distinguishing between the descriptions and not necessarily for describing a sequential or chronological order.

In addition, in the embodiment of the present application, "terminal device", "mobile phone", and the like may be used in combination, that is, various devices that may be used to implement the embodiment of the present application are referred to; in the embodiments of the present application, "application" and "application program" may also be mixed, and both refer to a program or a client having a certain service providing capability, that is, the application and the client may also be mixed, for example, the browser client and the game client may also be referred to as a browser application or a game application.

It should be understood that the hardware structure of the terminal device may be as shown in fig. 2a, and the software architecture may be as shown in fig. 2b, wherein a software program and/or a module corresponding to the software architecture in the terminal device may be stored in the memory 240, and the processor 230 may execute the software program and the process applied to execute the content recommendation method based on the user interest provided in the embodiment of the present application.

In order to facilitate understanding of the content recommendation method based on user interests provided by the present application, the steps included in the method provided by the present application are described below with reference to the schematic structural diagram shown in fig. 3. It should be noted that, in the embodiment of the present application, the division of the module is schematic, and is only one logic function division, and another division manner may be available in actual implementation. In addition, functional modules in the embodiments of the present application may be integrated into one processor, may exist separately and physically, or two or more modules are integrated into one unit. The integrated unit may be implemented in the form of hardware or a software program.

As shown in fig. 3, in the implementation of the present application, according to logic function division, a terminal device side may include a data acquisition module 301, a data calibration module 302, a privacy protection module 303, and a real-time recommendation module 306; the server side may include a data statistics module 304 and a data analysis module 305.

The data acquisition module 301 on the terminal device side is used for acquiring user operation behavior data, and the acquired user operation behavior data can be not only continuously preprocessed by the data calibration module 302, but also sent to the real-time recommendation module 306 for content recommendation of user interest. The privacy protection module 303 on the terminal device side is configured to further perform desensitization processing on the user operation behavior data after the data calibration module 302 performs preprocessing, and then send the desensitization processing to the data statistics module 304 on the server side.

The server-side data statistics module 304 is configured to perform statistics and summarization on desensitized user operation behavior data sent by one or more terminal devices (only one terminal device is shown in fig. 3 as an example, and if a plurality of terminal devices exist, the processing processes of other terminal devices are similar and are not described again), and then send the statistics and summarization to the data analysis module 305, and the data analysis module 305 performs training according to a large amount of summarized user operation behavior data to generate a topic interest table. The server-side data analysis module 305 may return the generated topic interest list to the terminal device. On one hand, the data analysis module 305 may not only send the generated topic interest table to the data calibration module 302, so that the data calibration module 302 may refer to the user operation behavior data for use in preprocessing, but on the other hand, the data analysis module 305 may also send the generated topic interest table to the real-time recommendation module 306, so that the real-time recommendation module 306 may perform real-time content recommendation based on the user interest in combination with the topic interest table. The server-side data statistics module 304 and the data analysis module 305 may also be integrated into a single module.

Based on the above description of the schematic structural diagram of fig. 3, the method provided by the present application can be mainly divided into several stages, which are specifically described as follows:

and in the stage 1, a data acquisition module 301 on the terminal equipment acquires user operation behavior data.

In order to ensure that the content recommended on the terminal device can more accurately track the preference and habit of the user, the content recommendation method based on the user interest is often implemented on massive user operation behavior data. The massive user operation behavior data are generally generated in the process of using the terminal equipment by the user. Generally, for any application included in the terminal device, the user operation performed from the time when the user starts the application to the time when the user closes the application may be taken as a complete set of user operation behaviors, for example, operation data included in the set of user operation behaviors may be recorded by a session (i.e., a time interval between a terminal user and a server providing the application service, and generally, a time elapsed between the user registering to enter the server providing the application service and logging off the server) object. One session object may include one or more pieces of operation data, and the number and types of the operation data are not limited in this application.

As an example, assuming that user A performs some search operation through a browser application, a session may be recorded from when user A opens the browser to when user A exits the browser. Specifically, in conjunction with the content shown in fig. 4, the user a opens the browser, which may be marked as the beginning of a session, as in interface 1 in fig. 4. During the period that the user a uses the browser, the terminal device may record all operations of the user a in this session through the data acquisition module 301, such as the interface 2 in fig. 4, which shows a home page interface of the browser, in which the user a may perform a search for a keyword through a search box, or perform a browsing of an information stream (the information stream is a word displayed on the home page interface of the browser, and the user a may also browse more information streams through a sliding operation up and down, which is not shown in fig. 4), or perform a browsing of entering a webpage through clicking the information stream on the home page interface of the browser (for example, the user a clicks a detailed interface of the information stream of "basketball association-home page" to browse more webpages about basketball consultation, which is not shown in fig. 4). Finally, if the terminal device detects that the user A exits the browser, the terminal device marks the end of the session; for example, the terminal device detects that the current display interface is changed to the main interface of the mobile phone (e.g., interface 3 in fig. 4), or the terminal device detects that the current display interface is switched to the display interface of another application program, or any other state that indicates that the current display interface of the mobile phone no longer stays on the browser.

Based on the operation behavior of the user, there are many possible application scenarios, and the user operation behavior data acquired by the data acquisition module 301 is characterized by multi-domain behavior. The multi-domain behavior indicates that a user performs operations on different display pages with different protocols, domain names, ports and the like, for example, if any two display pages included in the user operation behavior data adopt the same protocol, domain name, port and the like, the two display pages belong to the same domain; on the contrary, if any two display pages included in the user operation behavior data adopt different protocols, the two display pages belong to different domains, that is, the user operation behavior data is expressed as multi-domain behavior. For example, a user performs behaviors such as searching, information flow browsing, and web page browsing in a browser, where web pages corresponding to different information flows generally belong to different domains, so that user operation behavior data of the user in the browser generally has characteristics of a multi-domain behavior, that is, characteristics of a cross-domain behavior.

By collecting the user operation behavior data with the multi-domain behavior characteristics, after the massive user operation behavior data are subjected to statistical analysis, the interest topic tables in the multiple domains can be synchronized, and the user interests after the interest characteristics of the user operation behaviors in the multiple domains are integrated are obtained, so that more accurate content recommendation can be performed.

And 2, preprocessing the user operation behavior data acquired by the data acquisition module 301 by a data calibration module 302 on the terminal device.

For example, during each session, the sequence length of the user operation behavior data during each session may be inconsistent due to different operations of the user, where the sequence length of the user operation behavior data is determined according to the operation times of the user. For example, if the user performs a few user operation behaviors in the browser after opening the browser, the sequence length of the user operation behavior data collected during the session is shorter; if the user opens the browser and performs a plurality of user operation behaviors such as searching, information flow browsing, web browsing and the like in the browser, the sequence length of the user operation behavior data collected during the session is long. Table 1 is an example of user operation behavior data collected during a session, as follows:

TABLE 1

User ID	Key identification	Type of user operation behavior	Interest topic	Additional summary information (not necessarily filling)
					User A	NBA	Searching	Sports
User A	https://china.nba.com/	Web page browsing	Sports	NBA
					User A	Physical economy	Information flow browsing	Finance and economics
User A	New crown	Web page browsing	Current	Vaccine, pneumonia, and asymptomatic

In this embodiment, as one row of data in the user operation behavior data in table 1 above represents a group of sequences, and table 1 includes 4 rows of actual user operation behavior data, so the sequence length in table 1 may be considered as 4, similar tables in subsequent embodiments have the same definition, and repeated parts are not described again in the specific introduction process. The information types of the user operation behavior data shown in table 1 may include: the method comprises the following steps of obtaining information such as a user ID, a key identification used for embodying a user operation behavior, a user operation behavior type, an interest topic corresponding to the key identification, additional summary information (not necessarily filled) and the like. In addition, the user operation behavior data may further include more or less different information types than the above table 1, for example, application program identifiers, and the like, which is not limited in this application. The storage form of the user operation behavior data may be a table form or other forms, which is not limited in this application.

In the above example, the key identifier in table 1 may be obtained according to a user operation behavior, and the key identifier may be a search term, an information flow keyword, a web page address, and the like. For example, if the user operation behavior corresponds to a search operation and the search term is "NBA", the key identifier may be "NBA". Alternatively, if the user operation behavior corresponds to the information flow browsing and the information flow keyword is "basketball" (e.g., item 2 and item 3 information flows shown as item 2 in fig. 4), the key identifier may be "basketball". Or, if the user operation behavior corresponds to web page browsing, and the keyword containing the content in the web page is "entity economy", the key identifier may be the web page keyword "entity economy", or may also be the website of the currently browsed web page, and so on.

In a possible preprocessing scenario, the interest topic in table 1 may not be obtained from the user operation behavior (for example, when the user performs a search word operation, the terminal device may obtain a key identifier from the user operation behavior, but cannot determine the interest topic), and then the terminal device may determine the interest topic according to the key identifier in the user operation behavior and the topic interest table obtained from the server side. The topic interest table is used for indicating a mapping relationship between the key identifier and the interest topic, and is generated by the server side after performing statistical analysis based on the user operation behavior data received from one or more terminal devices and processed by the data calibration module 302 and the privacy protection module 303, and a specific generation manner is described in detail in the following embodiments, which is not detailed herein. Further, the topic interest table on the terminal device side may be periodically obtained from the server side and stored on the terminal device. The obtaining mode may be actively requested by the terminal device side, or may be issued periodically by the server, or may be actively issued after the server detects that the topic interest list is updated, and the implementation mode of obtaining the topic interest list from the server side by the terminal device is not limited in the present application.

Specifically, after the terminal device obtains the key identifier from the user operation behavior, the interest topic corresponding to at least one key identifier included in the user operation behavior at this time may be further determined according to the topic interest table. For example, if the key identifier acquired by the terminal device is "NBA", and the interest topic table includes a mapping relationship between "NBA" and the interest topic "sports", the terminal device may determine that the interest topic corresponding to the key identifier "NBA" is "sports" by querying the interest topic table, and store the interest topic and the key identifier together as the user operation behavior data, such as the data content shown in line 2 in table 1 above. In another possible preprocessing scene, in order to ensure the accuracy and the processing efficiency when modeling is performed on the basis of a large amount of user operation behavior data acquired during the session, the user operation behavior data acquired during the session are sampled according to a preset value during the implementation of the application, and then the user operation behavior data acquired during each session has a relatively fixed sequence length. The specific implementation is that the user operation behavior data with shorter sequence length is complemented by default user operation behavior, and the user operation behavior data with longer sequence length is randomly truncated and sampled. Therefore, the user operation behavior data with uniform behavior sequence length can be obtained, so that the problems that the user operation behavior data has short sequence length and can be understood as too little sample data to accurately analyze the user interest, and the user operation behavior data has long sequence length and can be understood as too much sample data to cause low processing efficiency due to relatively redundancy during analysis can be avoided.

As an example, the processing by the data calibration module 302 is described below in conjunction with tables 2 (including tables 2a, 2b) and 3 (including tables 3a, 3b) as follows:

TABLE 2a

Key identification	Subject of interest
		NBA	Sports
Physical economy	Finance and economics

Table 2a is an example of user operation behavior data collected during any session, and the sequence length of the user operation behavior data that can be obtained is short.

TABLE 2b

Key identification	Subject of interest
		NBA	Sports
Physical economy	Finance and economics
		“0”	“0”

As can be obtained from the above tables 2a and 2b, assuming that the data calibration module 302 performs sampling according to the preset value of the sequence length of the user operation behavior data being 3, when the sequence length of the user operation behavior data acquired by the data acquisition module 301 is short (it can also be understood that the sequence length is smaller than the preset value, for example, when the sequence length in table 2a is 2 and is smaller than the preset value 3), the predefined default user operation behavior data with the target length may be supplemented, so as to obtain the user operation behavior data with the sequence length of 3. The target length is an absolute value of a difference between the sequence length of the user operation behavior data and a preset value, for example, if the sequence length in table 2a is 2, and the preset value is 3, the target length is |2-3| 1, and then 1 sequence of user operation behavior data is supplemented in table 2 b. It should be noted that, in the above table 2b, a default key identifier is represented by "0", and a default interest topic mapped to the default key identifier is represented by "0". The default key identifier and the default interest topic may be preset, or determined according to a certain rule (e.g., according to a user operation behavior of the current hotspot), for example, the default key identifier may be "new crown", and the default interest topic is "hour administration".

TABLE 3a

Key identification	Interest topic
		NBA	Sports
Physical economy	Finance and economics
		Football game	Sports
https://china.nba.com/	Sports
		New crown	Current

Table 3a is an example of user operation behavior data collected during any session, and the sequence length of the user operation behavior data may be obtained to be longer.

TABLE 3b

Key identification	Interest topic
		NBA	Sports
Football game	Sports
		New crown	Current

As can be obtained from the above tables 3a and 3b, assuming that the data calibration module 302 performs sampling according to that the preset value of the sequence length of the user operation behavior data is 3, when the sequence length of the user operation behavior data acquired by the data acquisition module 301 is long (it can also be understood that the sequence length is greater than the preset value, for example, the sequence length in table 3a is 5 greater than the preset value 3), the target length may be randomly truncated, so as to obtain the user operation behavior data with the sequence length of 3. If the target length is an absolute value of a difference between the sequence length of the user operation behavior data and a preset value, for example, the sequence length in table 3a is 5, and the preset value is 3, and if the target length is |5-3| ═ 2, the user operation behavior data of the two sequences are truncated in table 3 b. Or, optionally, when the sequence length of the user operation behavior data is greater than the preset value of sampling, in addition to the implementation of randomly truncating the target length, the implementation of weighted sampling may be performed according to the behavior type of the user operation behavior, and then several groups of sequences included in the user operation behavior data with a larger weight are selected according to the preset value, for example, after the terminal device obtains the weight size of 5 groups of sequences included in table 3a according to the type of the user operation behavior, 3 groups of sequences with a larger weight are selected, such as 3 groups of sequences shown in table 3 b. Or, other implementation manners of sampling to obtain user operation behavior data with a sequence length of a preset value may also be adopted in implementation, which is not limited in the present application.

When the method is implemented, the user operation behavior data collected in the stage 1 can be preprocessed into a data structure with uniform sequence length through the processing of the stage 2, so that after various user operation behavior data generated in various scenes are preprocessed, desensitization processing can be continued through the privacy protection module 303 conveniently, and modeling of a content recommendation system based on user interest can be conveniently realized. The modeling for implementing the content recommendation system based on the user interests is specifically introduced in the subsequent embodiments, which are not repeated herein.

And 3, the privacy protection module 303 on the terminal device performs desensitization processing on the user operation behavior data preprocessed by the data calibration module 302 to obtain desensitized user operation behavior data.

When the application is implemented, desensitization processing on the user operation behavior data can be realized through one or a combination of the following modes, or desensitization processing can also be performed on the user operation behavior data through other possible modes, which is not limited in the application. Exemplary, include:

in the method 1, the terminal device performs random replacement processing on the user operation behavior data under the same interest topic by adopting a differential privacy algorithm. Specifically, for each user operation behavior included in the user operation behavior data corresponding to each session period, a certain probability remains unchanged, and a certain probability is randomly replaced. Optionally, the terminal device is implemented to keep the topic interest unchanged, and select a key identifier under the topic interest by randomly searching the topic interest table, and replace the original key identifier. Or, additional summary information in the user operation behavior data is regenerated. Following the example of the user operation behavior data in table 1, assuming that the privacy processing performed by the terminal device is to keep the subject interest "sports" in line 2 in table 1 unchanged, replace the key identification in line 2 with "basketball" from "NBA", and change the additional summary information in line 5 in table 1 from "vaccine, pneumonia, asymptomatic" to "wuhan city", the user operation behavior data after the privacy processing may be as shown in table 4 below:

TABLE 4

User ID	Key identification	Type of user operation behavior	Subject of interest	Additional summary information (not necessarily filling)
					User A	Basketball	Searching	Sports
User A	https://china.nba.com/	Web page browsing	Sports	NBA
					User A	Physical economy	Information flow browsing	Finance and economics
User A	New crown	Web page browsing	Current	Wuhan City

And 2, the terminal equipment strips the user privacy data in the user operation behavior data. Optionally, if the user operation behavior data collected by the terminal device includes user privacy data such as a user ID and a user operation behavior type, the privacy data may be stripped. Along with the example of the user operation behavior data in table 1, table 5 is an example of the user operation behavior data after stripping the user-related information, as follows:

TABLE 5

Key identification	Interest topic	Additional summary information (not necessarily filling)
			NBA	Sports
https://china.nba.com/	Sports	NBA
			Physical economy	Finance and economics
New crown	Current	Vaccine, pneumonia, asymptomatic

In the above example, if the terminal device performs desensitization processing on the user operation behavior data by the mode 1 and the mode 2, the processed user operation behavior data may be merged with the result processed by the two modes. Illustratively, table 6 is an example of the user operation behavior data after desensitization processing is performed on the user operation behavior data by way 1 and way 2, as follows:

TABLE 6

Key identification	Interest topic	Additional summary information (not necessarily filling)
			Basketball	Sports
https://china.nba.com/	Sports	NBA
			Physical economy	Finance and economics
New crown	Current	Wuhan City

As can be known from table 6 above, after the desensitization processing is further performed on the preprocessed user operation behavior data in stage 3, the obtained desensitization-processed user operation behavior data can achieve better protection of user privacy, and mainly aims to highlight the user interests of the current user operation behavior to obtain currently popular subject interests, hotspot key identifiers under each subject interest, and the like, so that a server side can generate a subject interest table or update the subject interest table conveniently, and timeliness and accuracy of content recommendation based on the user interests are improved.

And 4, the privacy protection module 303 on the terminal equipment uploads the processed desensitized user operation behavior data to the data statistics module 304 on the server side.

The server side may be connected to one or more terminal devices, so that the server side may obtain multiple sets of desensitized user operation behavior data uploaded by the privacy protection module 303 in the one or more terminal devices.

And 5, a data counting module 304 in the server performs summary counting on the desensitized user operation behavior data uploaded by the privacy protection module 303 in the received one or more terminal devices. And a data analysis module 305 in the server analyzes the desensitized user operation behavior data collected and counted by the data counting module 304 to generate a topic interest table.

It should be noted that, in the implementation of the present application, in consideration of the computing capability of the terminal device side, if the terminal device trains to generate the topic interest table, the terminal device needs to have a higher performance requirement, however, this implementation has the defects that the cost is higher and the recommendation efficiency cannot be improved well. Therefore, the training of obtaining the topic interest table according to the collected massive user operation behavior data can be generally realized on the server side. In addition, in order to protect the user privacy data, before uploading the user operation behavior data to the server, the terminal device performs desensitization processing on the sensitive data on the user operation behavior data, such as the preprocessing introduced in the phase 2 and the desensitization processing introduced in the phase 3 in the foregoing embodiment; and then, the terminal equipment sends the desensitized user operation behavior data obtained after the processing to a server to perform the operation of generating the topic interest list.

Illustratively, based on that each user operation behavior data includes mapping relationships between a plurality of key identifiers and interest topics, the data statistics module 304 may perform statistical analysis on a large amount of received desensitization-processed user operation behavior data, where the statistical analysis may include one or a combination of a corresponding relationship between a desensitization-processed user operation behavior and a key identifier, a corresponding relationship between desensitization-processed user operation behavior data and an interest topic, and a corresponding relationship between an interest topic and a key identifier.

1) The popular key identification searched or browsed by the current user can be obtained by analyzing the corresponding relation between the user operation behavior after desensitization processing and the key identification, so that the content recommendation of the terminal equipment can be realized according to the popular key identification, namely, the contents such as information streams, webpages and the like related to the popular key identification are mainly obtained for recommendation, and the content recommended by the terminal equipment is the content which is possibly interested by the user. For example, the statistical analysis result indicates that the number of times of the key identifier "NBA" included in the large amount of desensitized user operation behavior data is the largest, the terminal device determines that "NBA" is a popular key identifier searched or browsed by the current user, and in this scenario, the terminal device may obtain some content related to "NBA" for recommendation.

2) By analyzing the corresponding relation between the desensitized user operation behavior data and the interest topics, the popular interest topics searched or browsed by the current user can be obtained, and therefore content recommendation of the terminal device according to the popular interest topics can be achieved. For example, the statistical analysis result indicates that the maximum number of times of the interest topic "sports" is included in the large amount of desensitized user operation behavior data, the terminal device determines that "sports" is a popular interest topic searched or browsed by the current user, and in this scenario, the terminal device may obtain some content related to the "sports" interest topic for recommendation.

3) By analyzing the corresponding relation between the interest topic and the key identification, the hot key identification under each interest topic can be obtained, so that the terminal equipment can further recommend content according to the hot key identification under the scene of recommending the content according to the hot interest topic. For example, the statistical analysis result indicates that in a large amount of desensitized user operation behavior data, it is determined that key identifiers included under the interest topic "sports" are "NBA", "football", and the like, and the number of times of including "NBA" is the largest, the terminal device determines that a hot key identifier under the "sports" interest topic is "NBA", and when some contents related to the "sports" interest topic are acquired on the basis that the terminal device determines that the hot interest topic is "sports", some contents related to "NBA" may be acquired more for recommendation.

Taking the statistics of the corresponding relationship between the interest topic and the key identifier by the data statistics module 304 as an example, a specific embodiment is that after the data statistics module 304 receives a large amount of desensitized user operation behavior data uploaded by the privacy protection module 303 in one or more terminal devices, the mapping relationship between a plurality of key identifiers and topic interests is determined according to a plurality of sequences included in each desensitized user operation behavior data. The data statistics module 304 may determine a set of key identifiers included in each interest topic based on different interest topics as categories, so as to obtain a mapping relationship between each interest topic and the set of key identifiers included in the interest topic. For example, following the example in table 6, after the summary statistics of data statistics module 304, the mapping of table 7 is obtained as follows:

TABLE 7

Sports	Finance and economics	Current
			Basketball	Physical economy	New crown
https://china.nba.com/

Further, the data analysis module 305 may train the topic interest model after obtaining a statistical summary of the desensitized user operation behavior data. For example, the topic interest model may be constructed by using an algorithm such as an implicit dirichlet allocation (LDA) topic model algorithm, which is not limited in the present application.

The topic interest model may generate topic interest tables, user interest topic radar maps, and the like to determine user interests. The topic interest table may be in the form shown in table 7, each column represents an interest topic, and each column contains one or more key identifiers, where each column of interest topic contains key identifiers analyzed from massive desensitization-processed user operation behavior data. In addition, each column of interest topics may further be associated with a corresponding weight value, and the degree of interest of the user in the interest topic is represented by the weight value, where the weight value may be obtained by statistics and analysis of a large amount of desensitized user operation behavior data, for example, the weight value corresponding to an interest topic with a large number of occurrences in the desensitized user operation behavior data is larger. It is understood that the greater the weight value associated with the interest topic, the greater the interest level of the interest topic for most users.

Alternatively, fig. 5 is an exemplary diagram of a radar map of a user interest topic shown in this embodiment of the application, and it is assumed that the data analysis module 305 learns that the interest topic "sports" in a large amount of user operation behavior data after desensitization processing is the most interest topic searched and browsed by a user, "time", and "finance" are less, through analysis of the user operation behavior data after desensitization processing, and then generates a radar map that can reflect the interest degree of the user operation behavior after desensitization processing on the interest topic, as shown in fig. 5.

And 6, the data analysis module 305 in the server side sends the topic interest table to the data calibration module 302 in the terminal equipment side so as to realize content recommendation based on the user interest.

In an optional implementation manner, the topic interest table generated by the data analysis module 305 on the server side may be used by the data calibration module 302 on the terminal device side to pre-process the user operation behavior data, so as to obtain the pre-processed user operation behavior data. As described in the foregoing stage 2, the data calibration module 302 on the terminal device side may be implemented by combining with the topic interest table acquired from the server in the process of preprocessing the acquired desensitized user operation behavior data, so that the preprocessed desensitized user operation behavior data may include the key identifier and the interest topic, and further obtain more accurate desensitized user operation behavior data.

In another optional implementation manner, the topic interest table generated by the data analysis module 305 on the server side may also be used by the privacy protection module 303 on the terminal device side to process the preprocessed user operation behavior data through a differential privacy algorithm, so as to obtain the desensitized user operation behavior data. As described in the foregoing section 3, the privacy protection module 303 of the terminal device may randomly replace the content in the preprocessed user operation behavior data based on the topic interest table, so as to protect the privacy of the user operation behavior data.

In yet another alternative embodiment, the topic interest table generated by the data analysis module 305 on the server side may be used by the real-time recommendation module 306 on the terminal device side to perform real-time content recommendation, such as the content described in the following section 7, which will not be described in detail here.

And 7, determining user interest by a real-time recommendation module 306 on the terminal device according to the user operation behavior data acquired in real time by the data acquisition module 301 and a topic interest table acquired from the server side, and performing real-time recommendation according to the user interest.

For example, the real-time recommendation by the terminal device may include the following scenarios:

scenario 1, the user opens the application as a new user. In this scenario, the historical user interests of the user are not stored in the terminal device for the application.

In implementation, after the terminal device detects that the user enters the user operation behavior of the application program and before the terminal device does not detect that the user enters other user operation behaviors in the application program, content recommendation can be performed according to the weight value of the interest topic contained in the topic interest table. For example, if the terminal device detects that the user opens the browser, and before the user operation behavior of the user in the browser is not detected, the interest topics of "sports", "politics", and "finance" are included in the topic interest table, the terminal device may obtain the relevant content of these interest topics from the server side, recommend through the information flow on the top page of the browser, and if the weight value of the interest topic is greater, the proportion of the recommended relevant content is higher. Illustratively, the topic interest table is exemplified by the following table 8 a:

TABLE 8a

Further, after the terminal device detects the user real-time operation behavior of the user in the application program, the user interest may be determined by combining the user real-time operation behavior and the topic interest table acquired from the server side, and then real-time recommendation may be performed according to the user interest. For example, the real-time operation behavior of the user is represented by that the user inputs a search word "dragon fruit" in a browser, and the terminal device generates the user interest of the user according to the key identifier "dragon fruit" and the interest topic "fruit" corresponding to the key identifier, together with the topic interest table (as shown in table 8b below).

In specific implementation, the terminal device may associate different weight values for the interest topic according to the type of the user operation behavior. For example, since the search operation can better reflect the personal interests of the user, a higher weight value can be assigned to the interest topic "fruit" so that the terminal device recommends a higher proportion of the related content of the "fruit". For another example, when the terminal device detects that a user clicks on a certain information stream in a process of browsing home page interface information streams, the terminal device determines that a key identifier included in the information stream is "lion seat" and a corresponding interest topic is "constellation", and then the key identifier may be added to the user interest. Since the information stream browsing operation is generally expressed as the instant interest of the user, a lower weight can be assigned to the "constellation" of interest topic, so that the terminal device recommends a lower proportion of the "constellation" related content.

It should be noted that, as the number of operation behaviors of the user in the browser increases, the weight of the corresponding interest topic may be updated according to the number of operation behaviors. For example, if the user browses the content related to the "constellation" for a plurality of times in the browser subsequently, the weight value assigned to the "constellation" may be increased as the browsing times of the user increase. The user interests may be embodied in a personalized topic interest table form, as shown in the following table 8 b:

TABLE 8b

Fruit	Sports	Finance and economics	Current	Constellation
					Dragon fruit	Basketball	Physical economy	New crown	Lion seat
	https://china.nba.com/

The topic interest table shown in table 8b above is only one possible example and is not intended to limit the embodiment of the user interest. In the implementation of the present application, the interested topics shown in table 8b may also be sorted from left to right according to the magnitude of the weighted value, and the key identifiers included in each interested topic may also be sorted from top to bottom according to the magnitude of the weighted value. That is, it can be understood that the weight value associated with the interesting subject "fruit" in table 8b is currently the largest, so the percentage of the related content recommendations displayed on the terminal device as "fruit" is the highest.

In addition, after the terminal device detects that the user leaves the application program this time, the user interest obtained according to the real-time operation behavior of the user this time can be stored as the historical user interest of the user, so as to be used as a reference for determining the user interest when the user enters the application program next time.

Scenario 2, the user opens the application as an old user. In this scenario, the terminal device typically stores historical user interests of the user for the application. It should be noted that, if the terminal device detects that the user is an old user but does not store the historical user interests of the user, the user interests may also be determined according to the embodiment described in the foregoing scenario 1.

In implementation, after the terminal device detects an instruction of starting an application program by a user and before other user operation behaviors of the user in a browser are not detected, content recommendation can be performed by combining historical user interests and a topic interest list acquired from a server side. At this time, the user interests determined by the terminal device may be determined by the personalized topic interest table as shown in table 8b, for example, "fruit", "constellation" in table 8b is historical user interests, and "sports", "financial", "political" in table 8b is obtained from the topic interest table acquired from the server side.

Further, after the terminal device detects the user real-time operation behavior of the user in the application program, the user interest may be updated in combination with the user real-time operation behavior and the personalized topic interest table, and then real-time recommendation may be performed according to the updated user interest. For example, the terminal device detects from the real-time operation behavior of the user that the browsing times of the user for the subject interest "constellation" are significantly increased, and the related associated identifiers of browsing include "Capricorn" and "constellation tendency", the terminal device increases the weight value assigned to the "constellation" and updates the key identifier under the "constellation". The personalized topic interest table corresponding to the updated user interest may be as shown in the following table 8 c:

TABLE 8c

Constellation	Fruit	Sports	Finance and economics	Current
					Lion base	Dragon fruit	Basketball	Physical economy	New crown
Capricorn		https://china.nba.com/
					Constellation fortune

It should be noted that, after the terminal device determines the user interest, the recommended related content and the manner of acquiring the related content are not limited. For example, the terminal device may search for recommended content corresponding to the user interest from locally cached content, or may acquire recommended content corresponding to the user interest from a content providing server that provides recommended content corresponding to the user interest.

Through the content recommendation method provided by the application, the problem of cold start of the information flow of the application program on the terminal equipment can be avoided. The problem of cold start of information flow of an application program generally occurs in a scene that a user opens the application program for the first time, and since historical user interests are not stored in terminal equipment, the terminal equipment cannot recommend the information flow. By the method, the terminal equipment can determine the real-time interest of the user according to the topic interest list acquired from the server side and the real-time operation behavior of the user, so that content recommendation can be performed according to the real-time interest of the user of the terminal equipment.

And when the method is implemented, the real-time interest of the user is generated at the terminal equipment side, compared with the prior art that after the user interest is generated at the server side according to the big data, basically consistent interest content is recommended to different users directly according to the user interest generated at the server side, the content recommendation method based on the user interest can update the user interest in time according to the real-time operation of the user, and therefore the recommended content can better reflect the content in which the user is interested.

In order to better understand the overall flow of the method provided by the present application, the method provided by the present application is further described below with reference to fig. 6. The method provided by the application can be mainly divided into two parts: a sampling part and a recommendation part. The terminal equipment can collect each operation of the user to obtain user operation behavior data. And then, the terminal equipment processes the user operation behavior data and then sends the user operation behavior data to the server side, so that the server side generates a theme interest table according to the collected large amount of desensitized user operation behavior data. And the server can also send the topic interest list to the terminal equipment so as to realize content recommendation by the terminal equipment according to the topic interest list. Fig. 6 is a schematic flow chart of content recommendation based on user interests according to an embodiment of the present application, including the following steps:

a sampling part

S601, the terminal device detects an instruction of a user for starting a target application program, and starts the target application program. Optionally, the user may enter the target application program by clicking an application program icon on a main interface of the terminal device, or the user may wake up the target application program through voice, or the user may enter the target application program through a shortcut of the target application program included in any display interface of the terminal device, which is not limited in the present application. For example, assuming that the target application is a browser, the terminal device detects a user operation behavior of clicking a browser icon by a user, and may refer to content shown in fig. 7 as 1; or the terminal device receives a wakeup word that the user calls the browser, similar to "open the browser", or the terminal device receives a shortcut entry identifier of the browser included in the pull-down interface clicked by the user, and the like, and may execute the operation of opening the browser with reference to the content shown in fig. 2 in fig. 7.

S602, the terminal equipment collects at least one operation data executed by the user on the target application program. Illustratively, as shown in fig. 4, after the terminal device enters the target application program, the user operation behavior is collected in real time.

S603, when the terminal device detects that the user exits the instruction of the target application program, closing the target application program. For example, the user quitting the target application program may be the user closing the display interface of the target application program and returning to the main display interface of the terminal device; or the user can close the running of the target application program in a background cleaning mode; or may be a forced exit of the target application program due to no response of the program, and the like, which is not limited in this application.

And S604, the terminal equipment stores at least one operation data of the processing procedures from S601 to S603 as a group of user operation behavior data.

S605, the terminal equipment preprocesses the user operation behavior data to obtain the preprocessed user operation behavior data with the sequence value of a preset value. For example, due to multiple possibilities of user operation behaviors, each set of user operation behavior data may have different sequence lengths, and the terminal device may sample the user operation behavior data of different sequence lengths based on a preset value. Specifically, if the sequence length of the user operation behavior data is smaller than the preset value, a default user operation behavior is supplemented to the user operation behavior data, wherein the default user operation behavior may be obtained from a topic interest table or predefined, and the user operation behavior data with the specified sequence length is obtained. And if the sequence length of the user operation behavior data is larger than the preset value, performing random truncation sampling processing on the user operation behavior data to obtain the user operation behavior data with the specified sequence length. Therefore, by preprocessing the sequence length of the user operation behavior data, the problems that the user operation behavior cannot be comprehensively reflected due to the short sequence of the user operation behavior data and the sample data is too large due to the long sequence of the user operation behavior data can be better avoided. It should be noted that the preset value may be self-defined by the terminal device, or obtained by summarizing based on historical experience, or determined according to other rules, which is not limited in the present application.

And S606, the terminal equipment performs desensitization processing on the preprocessed user operation behavior data to obtain the desensitized user operation behavior data. In one possible design, the terminal device may randomly replace the content included in the preprocessed user operation behavior data according to the topic interest table, so as to perform certain fuzzy processing on the interest characteristics of the user operation behavior. In another possible design, the terminal device may further perform user information stripping on the user operation behavior data after the fuzzy processing, so as to obtain the user operation behavior data after the desensitization processing, thereby avoiding revealing the privacy of the user.

The topic interest table involved in the implementation processes of S605 and S606 may be stored after the terminal device acquires the topic interest table from the server side, so that S6050 is located before S605 and S606, but the execution sequence between S6050 and S601 to S604 is not limited.

S6050, the terminal equipment obtains the topic interest table generated by the server side. Optionally, the server side may automatically send the topic interest list to the terminal device side in real time, periodically, or after the topic interest list is updated. Or, the terminal device may also send request information to the server side, and the server receives the request information and then issues the latest topic interest list to the terminal device.

And S607, the terminal equipment uploads the desensitized user operation behavior data to a server side. In consideration of the fact that the computing capacity of the terminal equipment side is limited, the server can conduct training of the topic interest model based on more user operation behavior data, and the training is facilitated to obtain a more accurate and comprehensive topic interest table. When the application is implemented, the desensitized user operation behavior data obtained after the terminal device performs the processing such as S605 and S606 on the user operation behavior data is uploaded to the server. According to the implementation method provided by the application, the terminal equipment side processes the collected user operation behavior data to obtain the desensitized user operation behavior data, and then uploads the desensitized user operation behavior data to the server side, so that the server side cannot collect the privacy data of the user, and the safety of the user operation behavior data can be improved.

And S608, the server performs statistics and summarization on the desensitized user operation behavior data uploaded by one or more terminal devices. In which, the server may be connected to one or more terminal devices, only one terminal device is illustrated in fig. 6 as an example, and the interaction between other terminal devices and the server is similar. Illustratively, by means of statistical summarization of the operation behavior data of the user after desensitization processing, the currently popular interest topics and the key identifications contained in each interest topic can be obtained.

And S609, the server generates or updates the topic interest table based on the user operation behavior data after desensitization processing after statistics summarization. For example, the server may use the user operation behavior data after statistics summary as a training sample to perform unsupervised learning on the user interest, thereby obtaining a topic interest table that may reflect a mapping relationship between an interest topic and a key identifier. The topic interest table may include a plurality of interest topics and a key identifier included in each interest topic. In addition, the topic interest table may further associate a weight value of each interest topic and a weight value of each key identifier included in each interest topic, and reflect the interest topic and the key identifier that are interested by most users through the size of the weight value. For example, if the weighted value of the interest topic "sports" contained in the topic interest table is the largest, it indicates that most users are interested in the sports topic; further, the key identifier "basketball" contained in the "sports" topic has the largest weight value, which indicates that most users are more interested in the keywords of basketball at present.

S610, the server sends the topic interest list to the one or more terminal devices.

Second, recommendation part

S611, the terminal device detects an instruction of starting the target application program by the user, and starts the target application program. For example, as shown in fig. 7, the terminal device detects a user operation behavior in which the user opens the browser again. It should be noted that, for more clear understanding of the present application, S601 to S603 and S611 to S613 are respectively used to represent processing performed by the terminal device in response to the user operation in two different scenarios, where S601 to S603 can also implement content recommendation according to the user interest while performing real-time collection of the user operation behavior; s611 to S6113 can also collect the user operation behaviors in real time while carrying out content recommendation according to the user interests.

And S612, the terminal equipment determines the user interest. For example, the terminal device may determine the user interest and may include the following possible scenarios:

scene A: the user opens the browser for the first time. In this scenario, since there is no historical user interest in the browser, the terminal device may recommend, according to the topic interest table, an information stream in the first page interface of the browser for the user, that is, display the information stream as the first recommendation interface. For example, according to the foregoing table 7, the interest topics acquired by the terminal device from the server side include "sports", "finance", and "political affairs", where the key identifier included in the "sports" includes basketball, https:// china.nba.com/, "finance" includes the key identifier included in the "finance" includes entity economy, and the key identifier included in the "political affairs" includes new crown, and when the user opens the browser for the first time, the browser home page interface may be as shown in fig. 8 a. According to the content shown in fig. 8a, when the user opens the browser for the first time, the content displayed on the browser home page interface includes the recommended information flow of the basketball association-home page related to the key identifier "basketball", the entry of the NBA chinese official website related to "https:// china. Furthermore, by sliding the browser's home page interface downwards by the user, it is also possible to browse recommended content (not shown in fig. 8 a) that is more related to the interest topic contained in the topic interest table.

Scene B: the user does not open the browser for the first time and does not perform the user operation. In such a scenario, historical user interests may be stored in the browser, and the terminal device may recommend information streams in a browser first page interface for the user according to the historical user interests and the topic interest table. For example, in combination with the topic interest table obtained in table 8b, besides "sports", "finance", and "politics" included in the scene a, the topic interest table may also include historical user interests "fruit", "constellation", and in this case, when the user opens the browser, the browser home page interface may be as shown in fig. 8 b.

Scene C: the user does not open the browser for the first time and does some user action. In such a scenario, the terminal device may recommend the information flow in the browser first page interface for the user according to the real-time operation behavior of the user, the historical user interest and the topic interest table, that is, display the information flow as the first recommendation interface. For example, as shown in the foregoing table 8c, according to the updated personalized topic interest table obtained by the user real-time operation behavior, it may be obtained that the user is more interested in the topic interest "constellation", and then the proportion of the content related to the "constellation" in the content recommendation of the browser home page interface is increased, and then the browser home page interface that is refreshed by the terminal device according to the updated personalized topic interest table may be as shown in fig. 8 c. Because the updated personalized topic interest list indicates that the user is more interested in the constellation, the recommendation proportion of the content related to the constellation in the browser home page interface is increased and the content is positioned at a position in the browser home page interface, which is farther forward.

In the above example implementation, the second recommendation interface may be triggered and displayed after the terminal device detects an instruction of the user to refresh the first recommendation interface. For example, in the application, when the user slides down from the top of the terminal device, it indicates that the user wants to refresh the current interface, and at this time, the terminal device may display the recommended content corresponding to the updated user interest on the terminal device, or may understand that the second recommended interface is displayed.

It should be noted that the user interest is not specific to a certain interest, and may represent a set of multiple interest topics, and each interest topic is associated with a weight value, where the weight value is used to reflect the degree of interest of the user, and a larger weight value of an interest topic may represent a larger interest of the user in the interest topic.

And S613, the terminal equipment carries out content recommendation according to the user interest. Optionally, the terminal device may obtain the relevant content corresponding to the user interest from the local cache, or may also send an obtaining request to the server, so as to obtain the relevant content corresponding to the user interest from the server, and the like.

Based on the same technical concept, fig. 9 shows a terminal device 900 provided in this embodiment of the present application. The terminal device 900 includes one or more processors 901; one or more memories 902; a communications interface 903, and one or more computer programs 904, which may be connected by one or more communications buses 905. The communication interface 903 is for enabling communications with other devices, such as a terminal device, for example, and may be a transceiver. Wherein the one or more computer programs 904 are stored in the memory 902 and configured to be executed by the one or more processors 901, the one or more computer programs 904 comprising instructions operable to perform steps comprising:

collecting a plurality of user operation behavior data input by a user when the user uses a target application program for one time or a plurality of times in a set time length; desensitizing the collected user operation behavior data, wherein the desensitizing treatment is to filter out privacy data related to the user in the user operation behavior data; and sending the desensitized plurality of user operation behavior data to a server so that the server analyzes the desensitized plurality of user operation behavior data to obtain a subject interest list of the user using the target application program.

For example, the collecting of a plurality of user operation behavior data input by a user when the user uses a target application program at one time is specifically implemented to start the target application program when an instruction for starting the target application program by the user is detected; after the target application program is started, collecting at least one operation data executed by the user on the target application program; when an instruction that the user exits the target application program is detected, closing the target application program; and storing at least one operation data collected from the starting to closing process of the target application program as a group of user operation behavior data.

Optionally, the desensitization processing is performed on the collected multiple user operation behavior data, and is specifically implemented by randomly replacing, based on a differential privacy algorithm, user operation behavior data under the same interest topic for one or more user operation behavior data in the multiple user operation behavior data, where the interest topic is determined according to the topic interest table; and stripping the user information contained in the plurality of user operation behavior data.

In a possible embodiment, before the random replacement of the user operation behavior data under the same interest topic is performed based on the differential privacy algorithm, the sequence length of each user operation behavior data is determined; and performing truncation and compensation processing on the user operation behavior data according to a preset value to obtain the user operation behavior data with the appointed sequence length.

The user operation behavior data is subjected to truncation and compensation processing according to a preset value to obtain user operation behavior data with a specified sequence length, and the specific implementation is that if the sequence length of the user operation behavior data is smaller than the preset value, the user operation behavior data is supplemented with user operation behavior data which is defined in advance and has a target length, and the user operation behavior data with the specified sequence length is obtained; if the sequence length of the user operation behavior data is larger than the preset value, the target length of the user operation behavior data is cut off to obtain the user operation behavior data with the specified sequence length; and the target length is the absolute value of the difference value between the sequence length of the user operation behavior data and a preset value.

Based on the same technical concept, fig. 9 also illustrates a server 900 provided in the embodiments of the present application. The server 900 includes one or more processors 901; one or more memories 902; a communications interface 903, and one or more computer programs 904, which may be connected by one or more communications buses 905. The communication interface 903 is for enabling communications with other devices, such as a terminal device, for example, and may be a transceiver. Wherein the one or more computer programs 904 are stored in the memory 902 and configured to be executed by the one or more processors 901, the one or more computer programs 904 comprising instructions operable to perform steps comprising:

receiving a plurality of user operation behavior data after desensitization processing sent by one or more terminal devices; the desensitized user operation behavior data is obtained by acquiring a plurality of user operation behavior data input by a user when the user uses a target application program for one time or a plurality of times in a set time length by one or a plurality of terminal devices and desensitizing the acquired plurality of user operation behavior data; the desensitization processing is to filter out privacy data related to the user in the user operation behavior data; analyzing the desensitized multiple user operation behavior data to obtain a subject interest list of the user using the target application program; and sending the topic interest list to the one or more terminal devices.

Illustratively, the analyzing the multiple user operation behavior data after the desensitization processing to obtain the subject interest table of the target application used by the user is specifically implemented by inputting the multiple user operation behavior data after the desensitization processing into a pre-constructed subject interest model to perform unsupervised learning on the multiple user operation behavior data after the desensitization processing; and obtaining a topic interest table output by the pre-constructed topic interest model.

Based on the same technical concept, fig. 9 also illustrates a terminal device 900 provided in the embodiments of the present application. The terminal device 900 includes one or more processors 901; one or more memories 902; a communications interface 903, and one or more computer programs 904, which may be connected by one or more communications buses 905. The communication interface 903 is for enabling communications with other devices, such as a terminal device, for example, and may be a transceiver. Wherein the one or more computer programs 904 are stored in the memory 902 and configured to be executed by the one or more processors 901, the one or more computer programs 904 comprising instructions operable to perform steps comprising:

receiving a theme interest table sent by a server, wherein the theme interest table is obtained by analyzing the desensitized operation behavior data of a plurality of users by the server; the desensitized user operation behavior data is obtained by acquiring a plurality of user operation behavior data input by a user when the user uses a target application program for one time or a plurality of times in a set time length for one or a plurality of terminal devices and desensitizing the acquired plurality of user operation behavior data; the desensitization processing is to filter out the private data related to the user in the user operation behavior data; when an instruction of the user for starting the target application program is detected, starting the target application program and displaying a first recommendation interface, wherein the first recommendation interface comprises at least one item of recommendation content; the at least one item of recommended content is determined from the topic interest table.

Illustratively, the starting the target application and displaying the first recommendation interface is specifically implemented by starting the target application; displaying a first recommendation interface after the target application program is started; taking one or more interest topics contained in the topic interest table as user interests, acquiring at least one item of recommended content according to the user interests, and displaying the acquired at least one item of recommended content in the first recommendation interface; each interest topic has an associated weight value, and the larger the weight value associated with an interest topic is, the higher the proportion of related content containing the interest topic in the recommended content is.

In a possible embodiment, after the target application program is started and the first recommendation interface is displayed, one or more user operation behavior data input by a user when the user uses the target application program are received and collected; when an instruction of refreshing the first recommendation interface by the user is detected, displaying a second recommendation interface; and the recommended content contained in the second recommendation interface is determined according to the one or more user operation behavior data and the topic interest table.

In a possible design, the displaying the second recommendation interface is specifically implemented by determining one or more corresponding interest topics according to the one or more user operation behavior data, and assigning an associated weight value to each interest topic; taking one or more interest topics corresponding to the user operation behavior data and one or more interest topics contained in the topic interest table as user interests, acquiring at least one item of recommended content according to the user interests, and displaying the acquired at least one item of recommended content in the second recommendation interface; each interest topic included in the topic interest table has an associated weight value, and the larger the weight value associated with an interest topic is, the higher the proportion of related content of the interest topic included in the recommended content is.

In a possible design, the obtaining of the at least one item of recommended content according to the user interest is specifically implemented by searching recommended content corresponding to the user interest from local cache content; and/or acquiring recommended content corresponding to the user interest from a content providing server providing the recommended content corresponding to the user interest.

Through the above description of the embodiments, it is clear to those skilled in the art that, for convenience and simplicity of description, the foregoing division of the functional modules is merely used as an example, and in practical applications, the above function distribution may be completed by different functional modules according to needs, that is, the internal structure of the device may be divided into different functional modules to complete all or part of the above described functions. For the specific working processes of the system, the apparatus and the unit described above, reference may be made to the corresponding processes in the foregoing method embodiments, and details are not described here again.

Each functional unit in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.

The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the embodiments of the present application may be essentially implemented or make a contribution to the prior art, or all or part of the technical solutions may be implemented in the form of a software product stored in a storage medium and including several instructions for causing a computer device (which may be a personal computer, a server, or a network device) or a processor to execute all or part of the steps of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: flash memory, removable hard drive, read only memory, random access memory, magnetic or optical disk, and the like.

The above description is only a specific implementation of the embodiments of the present application, but the scope of the embodiments of the present application is not limited thereto, and any changes or substitutions within the technical scope disclosed in the embodiments of the present application should be covered by the scope of the embodiments of the present application. Therefore, the protection scope of the embodiments of the present application shall be subject to the protection scope of the claims.

Claims

1. A content recommendation method based on user interests is applied to terminal equipment and comprises the following steps:

collecting a plurality of user operation behavior data input by a user when the user uses a target application program for one time or a plurality of times in a set time length;

desensitizing the collected user operation behavior data, wherein the desensitizing treatment is to filter out privacy data related to the user in the user operation behavior data;

and sending the desensitized multiple user operation behavior data to a server so that the server analyzes the desensitized multiple user operation behavior data to obtain a theme interest list of the user using the target application program.

2. The method of claim 1, wherein collecting a plurality of user operation behavior data input by a user when the user uses the target application program once comprises:

when an instruction of the user for starting the target application program is detected, starting the target application program;

after the target application program is started, acquiring at least one operation data executed by the user on the target application program;

closing the target application program when detecting the instruction of the user for exiting the target application program;

and storing at least one operation data collected from the starting to closing process of the target application program as a group of user operation behavior data.

3. The method according to claim 1 or 2, wherein desensitizing the collected plurality of user operational behavior data comprises:

randomly replacing the user operation behavior data under the same interest topic based on a differential privacy algorithm aiming at one or more user operation behavior data in the plurality of user operation behavior data, wherein the interest topic is determined according to the topic interest table;

and stripping the user information contained in the plurality of user operation behavior data.

4. The method according to claim 3, wherein before the random replacement of the user operation behavior data under the same interest topic based on the differential privacy algorithm, the method further comprises:

determining the sequence length of each user operation behavior data;

and performing truncation and compensation processing on the user operation behavior data according to a preset value to obtain the user operation behavior data with the appointed sequence length.

5. The method of claim 4, wherein truncating and compensating the user operation behavior data according to a preset value to obtain user operation behavior data with a specified sequence length, comprises:

if the sequence length of the user operation behavior data is smaller than the preset value, supplementing the user operation behavior data with the predefined user operation behavior data with the target length to obtain the user operation behavior data with the specified sequence length;

if the sequence length of the user operation behavior data is larger than the preset value, the target length of the user operation behavior data is cut off to obtain the user operation behavior data with the specified sequence length;

and the target length is the absolute value of the difference value between the sequence length of the user operation behavior data and a preset value.

6. A content recommendation method based on user interests is applied to a server and comprises the following steps:

receiving a plurality of user operation behavior data after desensitization processing sent by one or more terminal devices; the desensitized user operation behavior data is obtained by acquiring a plurality of user operation behavior data input by a user when the user uses a target application program for one time or a plurality of times in a set time length by one or a plurality of terminal devices and desensitizing the acquired plurality of user operation behavior data; the desensitization processing is to filter out privacy data related to the user in the user operation behavior data;

analyzing the desensitized multiple user operation behavior data to obtain a subject interest list of the user using the target application program;

and sending the topic interest list to the one or more terminal devices.

7. The method according to claim 6, wherein the analyzing the desensitization-processed plurality of user operation behavior data to obtain a topic interest list of the user using the target application program comprises:

inputting the plurality of desensitized user operation behavior data into a pre-constructed subject interest model so as to perform unsupervised learning on the desensitized user operation behavior data;

and obtaining a topic interest table output by the pre-constructed topic interest model.

8. A content recommendation method based on user interests is applied to terminal equipment and comprises the following steps:

receiving a theme interest table sent by a server, wherein the theme interest table is obtained by analyzing the desensitized operation behavior data of a plurality of users by the server; the desensitized user operation behavior data is obtained by acquiring a plurality of user operation behavior data input by a user when the user uses a target application program for one time or a plurality of times in a set time length by one or a plurality of terminal devices and desensitizing the acquired plurality of user operation behavior data; the desensitization processing is to filter out privacy data related to the user in the user operation behavior data;

when an instruction of the user for starting the target application program is detected, starting the target application program and displaying a first recommendation interface, wherein the first recommendation interface comprises at least one item of recommendation content; the at least one item of recommended content is determined from the topic interest table.

9. The method of claim 8, wherein the launching the target application and displaying a first recommendation interface comprises:

starting the target application program;

displaying a first recommendation interface after the target application program is started;

taking one or more interest topics contained in the topic interest table as user interests, acquiring at least one item of recommended content according to the user interests, and displaying the acquired at least one item of recommended content in the first recommendation interface; each interest topic has an associated weight value, and the larger the weight value associated with an interest topic is, the higher the proportion of related content containing the interest topic in the recommended content is.

10. The method of claim 8, wherein after the launching of the target application and the displaying of the first recommendation interface, the method further comprises:

receiving and collecting one or more user operation behavior data input by a user when the user uses the target application program;

when an instruction of refreshing the first recommendation interface by the user is detected, displaying a second recommendation interface; and the recommended content contained in the second recommendation interface is determined according to the one or more user operation behavior data and the topic interest table.

11. The method of claim 10, wherein displaying the second recommendation interface comprises:

determining one or more corresponding interest topics according to the one or more user operation behavior data, and distributing associated weight values for the interest topics;

taking one or more interest topics corresponding to the user operation behavior data and one or more interest topics contained in the topic interest table as user interests, acquiring at least one item of recommended content according to the user interests, and displaying the acquired at least one item of recommended content in the second recommendation interface; each interest topic included in the topic interest table has an associated weight value, and the larger the weight value associated with an interest topic is, the higher the proportion of related content of the interest topic included in the recommended content is.

12. The method according to claim 9 or 11, wherein the obtaining at least one item of the recommended content according to the user interest comprises:

searching recommended content corresponding to the user interest from local cache content; and/or the presence of a gas in the atmosphere,

and acquiring recommended content corresponding to the user interest from a content providing server providing the recommended content corresponding to the user interest.

13. A terminal device, comprising: one or more processors; one or more memories;

the one or more memories for storing one or more computer programs and data information; wherein the one or more computer programs comprise instructions;

the instructions, when executed by the one or more processors, cause the terminal device to perform the method of any of claims 1-5 or perform the method of any of claims 8-12.

14. A server, comprising: one or more processors; one or more memories;

the instructions, when executed by the one or more processors, cause the server to perform the method of claim 6 or 7.

15. A communication system comprising a terminal device according to claim 13 and a server according to claim 14.

16. A computer-readable storage medium comprising a computer program or instructions which, when run on a computer, causes the method of any one of claims 1-5 to be performed, or the method of claim 6 or 7 to be performed, or the method of any one of claims 8-12 to be performed.

17. A graphical user interface on a terminal device, the terminal device having a display screen, one or more memories, and one or more processors to execute one or more computer programs stored in the one or more memories, the graphical user interface comprising a graphical user interface displayed when the terminal device performs the method of any of claims 1-5 or performs the method of any of claims 8-12.