WO2023061276A1

WO2023061276A1 - Data recommendation method and apparatus, electronic device, and storage medium

Info

Publication number: WO2023061276A1
Application number: PCT/CN2022/124028
Authority: WO
Inventors: 姚利军
Original assignee: 维沃移动通信有限公司
Priority date: 2021-10-11
Filing date: 2022-10-09
Publication date: 2023-04-20
Also published as: CN113869063A

Abstract

The present application provides a data recommendation method and apparatus, an electronic device, and a storage medium. The method comprises: when a target page is displayed, determining a word vector corresponding to the target page; inputting the word vector into a deep learning model, and determining target data matching the word vector; and displaying the target data.

Description

Data recommendation method, device, electronic device and storage medium

Cross References to Related Applications

This application claims priority to Chinese Patent Application No. 202111182460.7 filed in China on October 11, 2021, the entire contents of which are hereby incorporated by reference.

technical field

The present application belongs to the technical field of the Internet, and in particular relates to a data recommendation method, device, electronic equipment and storage medium.

Background technique

With the rapid development of the mobile Internet, more and more users choose to use electronic devices to handle online business. However, when the user submits business-related data online, for example, when the user submits business-related documents online, it is necessary to manually check all the data stored in the electronic device, and upload the data after locating the relevant data.

In the above process, users need to query business-related data through cumbersome operations, which reduces the convenience of data uploading.

Contents of the invention

The purpose of the embodiment of the present application is a data recommendation method, device, electronic device and storage medium, which can solve the problem that users need to locate data through cumbersome operations, thereby reducing the convenience of data uploading.

In the first aspect, the embodiment of the present application provides a data recommendation method, which includes:

In the case of displaying the target page, determine the word vector corresponding to the target page;

The word vector is input into the deep learning model, and the target data matched with the word vector is determined;

Display the target data.

In the second aspect, the embodiment of the present application provides a data recommendation device, which includes:

The first determining module is used to determine the word vector corresponding to the target page when the target page is displayed;

The second determination module is used to input the word vector into the deep learning model, and determine the target data matched with the word vector;

The first display module is used to display the target data.

In a third aspect, an embodiment of the present application provides an electronic device, the electronic device includes a processor, a memory, and a program or instruction stored in the memory and operable on the processor, and the program or instruction is The processor implements the steps of the method described in the first aspect when executed.

In a fourth aspect, an embodiment of the present application provides a readable storage medium, on which a program or an instruction is stored, and when the program or instruction is executed by a processor, the steps of the method described in the first aspect are implemented .

In the fifth aspect, the embodiment of the present application provides a chip, the chip includes a processor and a communication interface, the communication interface is coupled to the processor, and the processor is used to run programs or instructions, so as to implement the first aspect the method described.

In the embodiment of the present application, when the target page is displayed, the word vector corresponding to the target page is determined; the word vector is input into the deep learning model, and the target data matching the word vector is determined; and the target data is displayed. In this way, when the user submits business-related data online, the user does not need to locate the data through cumbersome operations, but directly displays the target data when the target page is displayed, reducing the user's operations of locating the data Steps to recommend relevant data to users, thereby improving the convenience of data upload.

Description of drawings

Fig. 1 is a schematic diagram of the deep learning model provided by the embodiment of the present application;

FIG. 2 is a schematic diagram of a file analysis model provided by an embodiment of the present application;

FIG. 3 is a schematic diagram of a user behavior model provided by an embodiment of the present application;

FIG. 4 is a flowchart of a data recommendation method provided in an embodiment of the present application;

Fig. 5 is one of the application scene diagrams of the data recommendation method provided by the embodiment of the present application;

FIG. 6 is the second application scenario diagram of the data recommendation method provided by the embodiment of the present application;

FIG. 7 is a schematic flowchart of a data recommendation method provided in an embodiment of the present application;

FIG. 8 is a structural diagram of a data recommendation device provided by an embodiment of the present application;

FIG. 9 is a structural diagram of an electronic device provided by an embodiment of the present application;

FIG. 10 is a hardware structural diagram of an electronic device provided by an embodiment of the present application.

Detailed ways

The following will clearly describe the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, but not all of them. All other embodiments obtained by persons of ordinary skill in the art based on the embodiments in this application belong to the protection scope of this application.

The terms "first", "second" and the like in the specification and claims of the present application are used to distinguish similar objects, and are not used to describe a specific sequence or sequence. It should be understood that the terms so used are interchangeable under appropriate circumstances such that the embodiments of the application can be practiced in sequences other than those illustrated or described herein, and that references to "first," "second," etc. distinguish Objects are generally of one type, and the number of objects is not limited. For example, there may be one or more first objects. In addition, "and/or" in the specification and claims means at least one of the connected objects, and the character "/" generally means that the related objects are an "or" relationship.

In order to better understand the solution provided by this application, first describe the following:

Please refer to Fig. 1, Fig. 1 is the schematic diagram of the deep learning model that the embodiment of the present application provides, as shown in Fig. 1, above-mentioned deep learning model can be deep semantic matching model (Deep Structured Semantic Models, DSSM) model, generally speaking, The DSSM model is obtained from two network models in a federated learning environment, where federated learning is a machine learning technique. The DSSM model is also called the twin-tower model, wherein one network model can be a file analysis model, and the other network model can be a user behavior model.

Please refer to Figure 2 and Figure 3, as shown in Figure 2, one of the network models in the DSSM model is the file analysis model. The training process of the file analysis model is: obtain the file type and keywords of the training file, convert the above file type and keywords into word vectors, and the above word vectors are also called semantic embedding vectors, where the semantic embedding can be obtained by using word embedding technology vector. The semantic embedding vector is used as a training sample of the document analysis model, so that the document analysis model outputs a document feature vector. It should be understood that the document feature vector is associated with the training document.

Wherein, in the process of obtaining the file type and keywords of the training file, the security level of the training file can be determined based on the file type and keyword, and the privacy file with a higher security level can be determined.

Wherein, during the training process of the file analysis model, the training file and the corresponding file feature vector are stored in a preset mapping table, that is, the mapping table stores the mapping relationship between the file feature vector and the training file.

As shown in Figure 3, one of the network models in the DSSM model is the user behavior model. The training process of the user behavior model is: use relevant information that can represent user behavior as training information, such as the semantic embedding vector corresponding to the displayed page, user operation data, and user uploaded files, etc.; use these training information as the training of the user behavior model samples, so that the user behavior model outputs a behavior feature vector. It should be understood that the behavior feature vector is used to characterize the user's operation behavior, and the operation behavior includes the user's preference for file selection in a specific scenario.

During the training process of the DSSM model, the behavior feature vector and the file feature vector can be stored in the preset database, and based on the correlation between the behavior feature vector and the file feature vector, the storage location of the behavior feature vector and the file feature vector can be adjusted storage location.

For example, during a training process of the document analysis model, the input of the document analysis model is an identity document, and the output is the document feature vector corresponding to the identity document; during a training process of the user behavior model, if the input of the user behavior model includes Identity files and user operation data, the behavior feature vector output by the user behavior model and the file feature vector corresponding to the identity file can be stored in the database, and the storage location of the behavior feature vector and the storage of the file feature vector can be adjusted in the database position to reduce the Euclidean distance between the behavior feature vector and the document feature vector.

The data recommendation method provided by the embodiment of the present application will be described in detail below through specific embodiments and application scenarios with reference to the accompanying drawings.

The data involved in the embodiments of this application can be files, user information or account information, etc. For the purpose of clearly explaining the technical solution, the following uses the data as the implementation scenario of the file to explain the solution. It should be understood that the data is not described here. Specific limits.

Please refer to FIG. 4 . FIG. 4 is a flowchart of a data recommendation method provided by an embodiment of the present application. The data recommendation method provided in the embodiment of the present application includes the following steps:

S101. In the case of displaying a target page, determine a word vector corresponding to the target page.

The above-mentioned target page may be a data upload page, and the above-mentioned word vector is a semantic embedding vector. The data upload page may be of a certain website or of a certain application program. Wherein, for an application, the word vectors corresponding to the target pages associated with the application may be the same or different, which is not specifically limited here. Different applications can also have the same target page. For example, different banking applications have ID card upload pages. For different applications, the word vectors corresponding to the target pages of the application can be the same or different. Yes, no specific limitation is made here.

An optional implementation manner is that the target page can be preset, and if the currently displayed page is the preset page, then it is determined that the currently displayed page is the target page; another optional implementation manner is that the currently displayed page is Optical Character Recognition (OCR), if a specific field is detected, such as "file upload", "picture upload", etc., it is determined that the currently displayed page is the target page.

For ease of understanding, please refer to FIG. 5 , which is one of the application scenario diagrams of the data recommendation method provided by the embodiment of the present application. Figure 5 shows a scenario of a target page.

It should be understood that for specific technical solutions on how to determine the word vector corresponding to the target page, please refer to subsequent embodiments.

There can be one or more word vectors corresponding to the target page. Usually, there are multiple word vectors.

S102. Input the word vector into a deep learning model, and determine target data matching the word vector.

As mentioned above, the word vector corresponding to the target page can be input into the trained DSSM model to determine the target data matching the word vector. Wherein, the above-mentioned target data is the data recommended by the deep learning model with a high degree of matching with the above-mentioned word vector. For specific technical solutions, please refer to the subsequent embodiments.

S103. Display the target data.

In this step, after the target data is determined, the target data is displayed on the current page.

In this embodiment, when the target page is displayed, the word vector corresponding to the target page is determined; the word vector is input into the deep learning model, and the target data matching the word vector is determined; and the target data is displayed. In this way, when the user submits business-related data online, the user does not need to locate the data through cumbersome operations, but directly displays the target data when the target page is displayed, reducing the user's operations of locating the data Steps to recommend relevant data to users, thereby improving the convenience of data upload.

Optionally, the inputting the word vector into the deep learning model, and determining the target data matching the word vector includes:

The word vector is input into the deep learning model to obtain the first target feature vector;

determining at least one second target feature vector corresponding to the first target feature vector in the database;

The data corresponding to the second target feature vector is determined as the target data.

In this embodiment, the word vector can be input into the user behavior analysis model to obtain the first target feature vector, wherein the first target feature vector is a behavior feature vector representing user behavior.

In this embodiment, a database is preset, and the database stores the first feature vector and the second feature vector. After the first target feature vector is obtained, the nearest neighbor search algorithm or In another method, at least one second target feature vector corresponding to the first target feature vector is determined. Wherein, the above-mentioned second target feature vector is a file feature vector, and the second target feature vector is used to characterize the file.

As mentioned above, during the training process of the file analysis model, it is necessary to store the training files and the corresponding file feature vectors in a preset mapping table.

In this embodiment, after the second target feature vector is obtained, the second target feature vector is input into a preset mapping table to obtain corresponding target data, wherein the mapping table stores the relationship between the second target feature vector and the data Further, the data obtained through querying the mapping table is determined as the target data.

Optionally, the determining at least one second target feature vector corresponding to the first target feature vector in the database includes:

determining a storage location of the first target feature vector in the database;

Based on the storage location, using a nearest neighbor search algorithm to obtain the Euclidean distance between the first target feature vector and the second feature vector;

Determining a second eigenvector whose Euclidean distance is less than or equal to a preset threshold as the second target eigenvector.

In this embodiment, after the first target feature vector is obtained, the storage location of the first target feature vector in the database is determined. It should be understood that the database stores a first feature vector and a second feature vector, the first target feature vector is the first feature vector, and the second target feature vector is the second feature vector, wherein the first feature vector is a behavior representing user behavior A feature vector, the second feature vector is a file feature vector characterizing the file.

On the basis of the storage location, the nearest neighbor search algorithm is used to obtain the Euclidean distance between the first target feature vector and the second feature vector, and the above-mentioned Euclidean distance is used to characterize the distance between vectors. In other embodiments, other methods may also be used to calculate the Euclidean distance between vectors, which is not specifically limited here.

Further, the second feature vector whose Euclidean distance is less than or equal to the preset threshold is determined as the second target feature vector, where there may be multiple second target feature vectors.

In this embodiment, based on the Euclidean distance between the first target feature vector and the second feature vector, the second target feature vector associated with the first target feature vector is determined.

Optionally, the displaying the target data includes:

The target data corresponding to the second target feature vector is displayed in ascending order of the Euclidean distance.

It should be understood that one second target feature vector corresponds to one target data. In this embodiment, after the second target feature vectors are obtained, the target data corresponding to the second target feature vectors are displayed in ascending order of the Euclidean distances between each second target feature vector and the first target feature vector. That is to say, the Euclidean distance between the second target eigenvector corresponding to the first displayed target data and the first target eigenvector is the smallest, and the last displayed target data corresponds to the second target eigenvector and the first target eigenvector The Euclidean distance between them is the largest.

It should be understood that during the training process of the DSSM model, if the correlation between the behavior feature vector and the document feature vector is stronger, the Euclidean distance between the two feature vectors is shorter.

In this embodiment, the target data is sorted and displayed, and the target data corresponding to the second target feature vector that is strongly related to the first target feature vector is sorted first, so as to preferentially display the target data that is relevant to the target page, and the user There is no need to perform relevant operations to query the target data, thereby improving the convenience of data upload.

Optionally, the displaying the target data includes:

Obtain the data type and keywords corresponding to the target data;

When the data type is a preset type and/or the keyword includes a preset field, a reminder mark is displayed in a preset area of the target data.

In this embodiment, after the target data corresponding to the second target feature vector is determined, the data type and keywords corresponding to the target data are acquired.

An optional implementation manner is that, when the data type is a preset type, it is indicated that the content of the target data involves private information, the target data is determined as private data, and the security level of the target data is relatively high.

Another optional implementation manner is to determine the target data as private data when the keyword includes a preset field. Wherein, the above-mentioned keyword may be a keyword in the data name, or a keyword obtained after performing OCR processing on the data file.

Another optional implementation manner is to determine the target data as private data when the data type is a preset type and the keyword includes a preset field.

Further, a reminder mark is displayed in a preset area of the target data, and the reminder mark includes but not limited to text, image or graph.

In this embodiment, according to the data type and keywords of the target data, the target data related to private information is determined, and a reminder mark is displayed in the preset area of the private data to remind the user, so as to improve the security of the subsequent data upload process.

For ease of understanding, please refer to FIG. 6 . FIG. 6 is the second application scenario diagram of the data recommendation method provided by the embodiment of the present application. As shown in FIG. 6 , a dotted frame is displayed in the preset area of the target data, and the dotted frame is a reminder mark, reminding the user that the file is private data. In the scene shown in FIG. 6 , if the target data is a picture file, the picture is displayed directly; if the target data is a file, a preview image of the file is displayed.

Optionally, after displaying the target data, the method includes:

receiving user input on the target data;

If the application program corresponding to the target page is the target application program, display reminder information.

The above-mentioned input may be a user's touch input or sliding input or other types of input on the target data.

In this embodiment, after receiving the user's input of the target data, the application program corresponding to the target page and the security level corresponding to the target are detected. When the application corresponding to the target page is the target application and the target data is private data, a reminder message is displayed.

It should be understood that whether the application associated with the target page is a target application can be determined according to the category of the application. Optionally, the above target application is a non-government and enterprise application, such as a communication application or a film and television application.

It should be understood that the above reminder information may be text information or voice information. Exemplarily, the above reminder information may be a pop-up window displayed on the target page, and the content of the pop-up window is a text message of "currently a private file, continuing to select may lead to privacy disclosure".

In other embodiments, if the application program corresponding to the target page is not the target application program, or the target data is not private data, no reminder message will be displayed after receiving the user's input of the target data.

In this embodiment, when the application program corresponding to the target page is the target application program and the target data is private data, a reminder message is displayed, so as to prevent the user from revealing the private data, thereby improving the security of data uploading.

The technical solution for determining the word vector corresponding to the target page is described in detail below:

Optionally, the determining the word vector corresponding to the target page includes:

performing optical character recognition processing on the target page to obtain text information;

performing keyword extraction on the text information to obtain keyword information;

Word embedding processing is performed on the keyword information to obtain the word vector.

In this embodiment, when the target page is displayed, OCR processing is performed on the target page to identify text information corresponding to the target page, and the text information includes text displayed on the target page.

Keywords of the text information are extracted to obtain keyword information. Among them, you can use the Term Frequency–Inverse Document Frequency (TF-IDF) algorithm to extract keywords or use the document topic generation model (Latent Dirichlet Allocation, LDA) algorithm to extract keywords. Other methods, this embodiment No specific limitation is made here.

After the keyword information is obtained, word embedding processing is performed on the keyword information to obtain a word vector corresponding to the keyword information.

For easy understanding of the overall solution, please refer to FIG. 7 , which is a schematic flow chart of the data recommendation method provided by the embodiment of the present application.

As shown in Figure 7, in the case of displaying the file upload page, determine the semantic embedding vector corresponding to the file upload page, wherein the above-mentioned file upload page is the target page, and the above-mentioned semantic embedding vector is the word vector. Input the semantic embedding vector into the trained deep learning model to obtain the behavioral feature vector, wherein the above deep learning model is the deep learning model. The behavior feature vector is queried in a preset database to obtain a document feature vector, wherein the behavior feature vector is a first target feature vector, and the document feature vector is a second target feature vector. The file to be uploaded corresponding to the file feature vector is determined using a preset mapping table, and the file to be uploaded is data. Further, the files to be uploaded are displayed. When the input of the file to be uploaded is received, it is judged whether the application program corresponding to the file upload page is the target application program, and whether the file to be uploaded is a private file. If the application corresponding to the file upload page is the target application and the file to be uploaded is a private file, a reminder message is displayed. If the application program corresponding to the file upload page is not the target application program, and/or the file to be uploaded is not a private file, then the file to be uploaded is uploaded to the file upload page.

As shown in Figure 8, the data recommendation device 200 includes:

The first determining module 201 is configured to determine a word vector corresponding to the target page when the target page is displayed;

The second determination module 202 is used to input the word vector into the deep learning model, and determine the target data matched with the word vector;

The first display module 203 is configured to display the target data.

Optionally, the second determining module 202 is specifically configured to:

Optionally, the second determining module 202 is also specifically configured to:

Optionally, the first display module 203 is specifically configured to:

Optionally, the first display module 204 is further specifically configured to:

Obtain the data type and keywords corresponding to the target data;

Optionally, the data recommendation device 200 further includes:

a receiving module, configured to receive user input on the target data;

The second display module is configured to display reminder information when the application program corresponding to the target page is the target application program.

Optionally, the first determining module 202 is specifically configured to:

The data recommending device in the embodiment of the present application may be a device, or may be a component, an integrated circuit, or a chip in a terminal. The device may be a mobile electronic device or a non-mobile electronic device. Exemplary, the mobile electronic device can be a mobile phone, a tablet computer, a notebook computer, a handheld computer, a vehicle electronic device, a wearable device, an ultra-mobile personal computer (Ultra-Mobile Personal Computer, UMPC), a netbook or a personal digital assistant (Personal Digital Assistant). Assistant, PDA), etc., non-mobile electronic devices can be servers, network attached storage (Network Attached Storage, NAS), personal computer (Personal Computer, PC), television (TeleVision, TV), teller machine or self-service machine, etc., this application Examples are not specifically limited.

The data recommendation device in the embodiment of the present application may be a device with an operating system. The operating system may be an Android operating system, an ios operating system, or other possible operating systems, which are not specifically limited in the embodiments of the present application.

The data recommendation device provided in the embodiment of the present application can realize each process realized in the method embodiment in FIG. 4 , and details are not repeated here to avoid repetition.

Optionally, as shown in FIG. 9 , the embodiment of the present application further provides an electronic device 300, including a processor 301, a memory 302, and programs or instructions stored in the memory 302 and operable on the processor 301, When the program or instruction is executed by the processor 301, each process of the above-mentioned data recommendation method embodiment can be realized, and the same technical effect can be achieved. To avoid repetition, details are not repeated here.

It should be noted that the electronic devices in the embodiments of the present application include the above-mentioned mobile electronic devices and non-mobile electronic devices.

FIG. 10 is a schematic diagram of a hardware structure of an electronic device implementing an embodiment of the present application.

The electronic device 1000 includes, but is not limited to: a radio frequency unit 1001, a network module 1002, an audio output unit 1003, an input unit 1004, a sensor 1005, a display unit 1006, a user input unit 1007, an interface unit 1008, a memory 1009, and a processor 1010, etc. part.

Those skilled in the art can understand that the electronic device 1000 can also include a power supply (such as a battery) for supplying power to various components, and the power supply can be logically connected to the processor 1010 through the power management system, so that the management of charging, discharging, and function can be realized through the power management system. Consumption management and other functions. The structure of the electronic device shown in FIG. 10 does not constitute a limitation to the electronic device. The electronic device may include more or fewer components than shown in the figure, or combine certain components, or arrange different components, and details will not be repeated here. .

Wherein, the processor 1010 is further configured to determine a word vector corresponding to the target page when the target page is displayed;

The display unit 1006 is further configured to display the target data.

Wherein, the processor 1010 is further configured to input the word vector into the deep learning model to obtain the first target feature vector;

Determine at least one second target feature vector corresponding to the first target feature vector in the database.

Wherein, the processor 1010 is also configured to determine the storage location of the first target feature vector in the database;

Wherein, the display unit 1006 is further configured to display the target data corresponding to the second target feature vector in ascending order of the Euclidean distance.

Wherein, the processor 1010 is further configured to obtain the data type and keywords corresponding to the target data;

The display unit 1006 is further configured to display a reminder mark in a preset area of the target data when the data type is a preset type and/or the keyword includes a preset field.

Wherein, the user input unit 1007 is also used to receive the input of the target data from the user;

The display unit 1006 is further configured to display reminder information when the application program corresponding to the target page is the target application program.

Wherein, the processor 1010 is further configured to perform optical character recognition processing on the target page to obtain text information;

It should be understood that, in the embodiment of the present application, the input unit 1004 may include a graphics processor (Graphics Processing Unit, GPU) 10041 and a microphone 10042, and the graphics processor 10041 is used for the image capture device ( Such as the image data of the still picture or video obtained by the camera) for processing. The display unit 1006 may include a display panel 10061, and the display panel 10071 may be configured in the form of a liquid crystal display, an organic light emitting diode, or the like. The user input unit 1007 includes a touch panel 10071 and other input devices 10072 . The touch panel 10071 is also called a touch screen. The touch panel 10071 may include two parts, a touch detection device and a touch controller. Other input devices 10072 may include, but are not limited to, physical keyboards, function keys (such as volume control buttons, switch buttons, etc.), trackballs, mice, and joysticks, which will not be repeated here. The memory 1009 can be used to store software programs as well as various data, including but not limited to application programs and operating systems. Processor 1010 may integrate an application processor and a modem processor, wherein the application processor mainly processes an operating system, user interface, application program, etc., and the modem processor mainly processes wireless communication. It can be understood that the foregoing modem processor may not be integrated into the processor 1010 .

The embodiment of the present application also provides a readable storage medium, the readable storage medium stores a program or an instruction, and when the program or instruction is executed by a processor, each process of the above-mentioned data recommendation method embodiment is realized, and can achieve the same To avoid repetition, the technical effects will not be repeated here.

Wherein, the processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer readable storage medium, such as computer read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disk, etc.

The embodiment of the present application further provides a chip, the chip includes a processor and a communication interface, the communication interface is coupled to the processor, and the processor is used to run programs or instructions to implement the above data recommendation method embodiment Each process can achieve the same technical effect, so in order to avoid repetition, it will not be repeated here.

It should be understood that the chips mentioned in the embodiments of the present application may also be called system-on-chip, system-on-chip, system-on-a-chip, or system-on-a-chip.

It should be noted that, in this document, the term "comprising", "comprising" or any other variation thereof is intended to cover a non-exclusive inclusion such that a process, method, article or apparatus comprising a set of elements includes not only those elements, It also includes other elements not expressly listed, or elements inherent in the process, method, article, or device. Without further limitations, an element defined by the phrase "comprising a ..." does not preclude the presence of additional identical elements in the process, method, article, or apparatus comprising that element. In addition, it should be pointed out that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in reverse order according to the functions involved. Functions are performed, for example, the described methods may be performed in an order different from that described, and various steps may also be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

Through the description of the above embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus a necessary general-purpose hardware platform, and of course also by hardware, but in many cases the former is better implementation. Based on this understanding, the essence of the technical solution of this application or the part that contributes to related technologies can be embodied in the form of computer software products, which are stored in a storage medium (such as ROM/RAM, disk, CD) contains several instructions to enable a terminal (which may be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in various embodiments of the present application.

The embodiments of the present application have been described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementations. The above-mentioned specific implementations are only illustrative and not restrictive. Those of ordinary skill in the art will Under the inspiration of this application, without departing from the purpose of this application and the scope of protection of the claims, many forms can also be made, all of which belong to the protection of this application.

Claims

A data recommendation method, comprising:

In the case of displaying the target page, determine the word vector corresponding to the target page;

The word vector is input into the deep learning model, and the target data matched with the word vector is determined;

Display the target data.
The method according to claim 1, wherein said inputting said word vector into a deep learning model, determining target data matched with said word vector comprises:

The word vector is input into the deep learning model to obtain a first target feature vector, and the first target feature vector is used to represent user behavior;

Determining at least one second target feature vector corresponding to the first target feature vector in the database; the database stores a first feature vector and a second feature vector, and the first feature vector and the second feature vector are based on The deep learning model is generated;

The data corresponding to the second target feature vector is determined as the target data.
The method according to claim 2, wherein said determining at least one second target feature vector corresponding to the first target feature vector in the database comprises:

determining a storage location of the first target feature vector in the database;

Based on the storage location, using a nearest neighbor search algorithm to obtain the Euclidean distance between the first target feature vector and the second feature vector;

Determining a second eigenvector whose Euclidean distance is less than or equal to a preset threshold as the second target eigenvector.
The method of claim 3, wherein said displaying said target data comprises:

The target data corresponding to the second target feature vector is displayed in ascending order of the Euclidean distance.
The method of claim 1, wherein said displaying said target data comprises:

Obtain the data type and keywords corresponding to the target data;

When the data type is a preset type and/or the keyword includes a preset field, a reminder mark is displayed in a preset area of the target data.
The method according to claim 1, wherein, after displaying the target data, the method further comprises:

receiving user input on the target data;

If the application program corresponding to the target page is the target application program, display reminder information.
The method according to claim 1, wherein said determining the word vector corresponding to the target page comprises:

performing optical character recognition processing on the target page to obtain text information;

performing keyword extraction on the text information to obtain keyword information;

Word embedding processing is performed on the keyword information to obtain the word vector.
A data recommendation device, said device comprising:

A first determining module, configured to determine a word vector corresponding to the target page when the target page is displayed;

The second determination module is used to input the word vector into the deep learning model, and determine the target data matched with the word vector;

The first display module is used to display the target data.
The device according to claim 8, wherein the second determination module is specifically configured to:

The word vector is input into the deep learning model to obtain a first target feature vector, and the first target feature vector is used to represent user behavior;

Determining at least one second target feature vector corresponding to the first target feature vector in the database; the database stores a first feature vector and a second feature vector, and the first feature vector and the second feature vector are based on The deep learning model is generated;

The data corresponding to the second target feature vector is determined as the target data.
The device according to claim 9, wherein the second determination module is further specifically configured to:

determining a storage location of the first target feature vector in the database;

Based on the storage location, using a nearest neighbor search algorithm to obtain the Euclidean distance between the first target feature vector and the second feature vector;

Determining a second eigenvector whose Euclidean distance is less than or equal to a preset threshold as the second target eigenvector.
The device according to claim 10, wherein the first display module is specifically used for:

The target data corresponding to the second target feature vector is displayed in ascending order of the Euclidean distance.
The device according to claim 8, wherein the first display module is further specifically configured to:

Obtain the data type and keywords corresponding to the target data;

When the data type is a preset type and/or the keyword includes a preset field, a reminder mark is displayed in a preset area of the target data.
The device according to claim 8, wherein the device further comprises:

a receiving module, configured to receive user input on the target data;

The second display module is configured to display reminder information when the application program corresponding to the target page is the target application program.
The device according to claim 8, wherein the first determining module is specifically configured to:

performing optical character recognition processing on the target page to obtain text information;

performing keyword extraction on the text information to obtain keyword information;

Word embedding processing is performed on the keyword information to obtain the word vector.
An electronic device, comprising a processor, a memory, and a program or instruction stored in the memory and operable on the processor, wherein the program or instruction is executed by the processor to achieve claim 1 The steps of the data recommendation method described in any one of -7.
A readable storage medium, on which a program or instruction is stored, wherein, when the program or instruction is executed by a processor, the steps of the data recommendation method according to any one of claims 1-7 are realized .
A chip comprising a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run programs or instructions to implement the data recommendation method according to any one of claims 1-7 A step of.
A computer software product, wherein the computer software product is stored in a non-volatile storage medium, and when the computer software product is executed by at least one processor, the data according to any one of claims 1-7 is realized Steps in the recommended method.
An electronic device configured to execute the steps of the data recommendation method according to any one of claims 1-7.