CN111191112A

CN111191112A - Electronic reading data processing method, device and storage medium

Info

Publication number: CN111191112A
Application number: CN201910629021.2A
Authority: CN
Inventors: 苏丹
Original assignee: Tencent Technology Shenzhen Co Ltd
Current assignee: Tencent Technology Shenzhen Co Ltd
Priority date: 2019-07-12
Filing date: 2019-07-12
Publication date: 2020-05-22

Abstract

The embodiment of the application discloses a method, a device and a storage medium for processing electronic reading material data, wherein the method comprises the following steps: responding to a first operation triggered by an electronic reading collection column, and determining dimension information corresponding to the first operation as target dimension information in a plurality of dimension information corresponding to the electronic reading collection column; the electronic reading matter collecting column comprises electronic collecting reading matters; sending the target dimension information to a server; acquiring a data push list which is returned by the server and is associated with the target dimension information; the data push list comprises electronic recommended reading materials which are in association relation with the electronic collection reading materials. By the method and the device, the efficiency and the accuracy of data recommendation can be improved.

Description

Electronic reading data processing method, device and storage medium

Technical Field

The present application relates to the field of internet technologies, and in particular, to a method and an apparatus for processing electronic reading data, and a storage medium.

Background

When recommending electronic reading materials (for example, internet information such as books) to a target user, an existing electronic reading material recommending system can recommend data according to the click rate of the electronic reading materials counted in a book city (namely, an electronic reading material database), so that the electronic recommended reading materials received by each user are the same and are difficult to be attached to the interests and hobbies of each user, and the accuracy of data recommendation is low.

In addition, when a user needs to refer to other books associated with a certain book (e.g., book a), corresponding index information (e.g., a keyword manually entered into book a, etc.) needs to be manually entered into a book city to perform a related search, so as to search all books including the keyword of book a, since it cannot be guaranteed that the keyword imagined by the user belongs to the keyword of book a, performing a book search through the keyword intended by the user may possibly result in a low accuracy of data recommendation; in addition, in order to accurately find other books associated with book a, a plurality of index information needs to be manually entered, so that a long time of manual interaction needs to be consumed in the whole data recommendation process, and the efficiency of data recommendation is low.

Disclosure of Invention

The embodiment of the application provides an electronic reading data processing method, an electronic reading data processing device and a storage medium, and the efficiency and the accuracy of data recommendation can be improved.

One aspect of the embodiments of the present application provides a method for processing electronic reading data, where the method includes:

responding to a first operation triggered by an electronic reading collection column, and determining dimension information corresponding to the first operation as target dimension information in a plurality of dimension information corresponding to the electronic reading collection column; the electronic reading matter collecting column comprises electronic collecting reading matters;

sending the target dimension information to a server;

acquiring a data push list which is returned by the server and is associated with the target dimension information; the data push list comprises electronic recommended reading materials which are in association relation with the electronic collection reading materials.

acquiring target dimension information sent by a user terminal; the target dimension information is obtained by a user terminal responding to a first operation aiming at a plurality of dimension information corresponding to the electronic reading collection column; the electronic reading matter collecting column comprises electronic collecting reading matters;

acquiring an electronic recommended reading material having an association relation with the electronic collection reading material based on the target dimension information;

and generating a data push list associated with the target dimension information according to the electronic recommended reading materials, and returning the data push list to the user terminal.

Wherein the target dimension information comprises first dimension information;

the obtaining of the electronic recommended reading materials having an association relation with the electronic collection reading materials based on the target dimension information comprises:

pulling N pieces of data information matched with the electronic collection reading materials from a first recommendation database based on the first dimension information; n is a positive integer;

filtering the recommended S data information from the N data information, and taking the residual K data information after filtering as the electronic recommended reading materials which have an association relationship with the electronic collection reading materials; k ═ N-S and S is an integer greater than or equal to zero.

Wherein the method further comprises:

obtaining an article vector of each electronic reading according to a keyword carried by each electronic reading in an electronic reading database and user information associated with each electronic reading;

determining the electronic collection reading in the electronic reading database as a first article to obtain an article vector of the first article;

determining the electronic reading materials except the electronic collection reading materials in the electronic reading material database as second products to obtain the product vectors of the second products;

determining cosine similarity between the article vector of the first article and the article vector of the second article, taking the second article with the cosine similarity meeting recommendation conditions as data information matched with the electronic collection reading, and adding the data information matched with the electronic collection reading to the first recommendation database.

The obtaining of the item vector of each electronic reading according to the keyword carried by each electronic reading in the electronic reading database and the user information associated with each electronic reading includes:

acquiring all electronic readings in an electronic reading database and user behavior information related to all electronic readings;

analyzing the user behavior information to obtain user information associated with each electronic reading material, and obtaining a user sequence of each electronic reading material according to the user information associated with each electronic reading material;

extracting keywords carried by each electronic reading material, and obtaining a keyword sequence of each electronic reading material according to the keywords of each electronic reading material;

and constructing an article vector of each electronic reading material based on the user sequence of each electronic reading material and the keyword sequence of each electronic reading material.

The method for pulling N pieces of data information matched with the electronic collection reading materials from a first recommendation database based on the first dimension information comprises the following steps:

obtaining a first click model associated with the first item based on the first dimension information;

determining the data information in the first recommendation database as M candidate data information; m is the total amount of data information in the recommendation database;

determining predicted probability values for the M candidate data information based on the first click model;

sequencing the predicted probability values of the M pieces of candidate data information, and selecting N pieces of data information matched with the electronic collection reading materials from the sequenced M pieces of candidate data information on the basis of the predicted probability values of the sequenced M pieces of candidate data information; n is less than or equal to M.

Wherein the target dimension information comprises second dimension information;

acquiring the classification category of the electronic collection reading material based on the second dimension information, and pulling X data information associated with the classification category of the electronic collection reading material from an electronic reading material database; x is a positive integer;

filtering the recommended Z data information from the X data information to obtain the residual Y data information after filtering; y ═ X-Z) and Z is an integer greater than or equal to zero;

and acquiring the click confidence of the Y data information, and determining the electronic recommended reading materials which have an association relation with the electronic collection reading materials from the Y data information according to the click confidence of the Y data information.

The obtaining of the click confidence of the Y data information and the determining of the electronic recommended reading material having an association relationship with the electronic collection reading material from the Y data information according to the click confidence of the Y data information include:

acquiring a second click model from a second recommendation database, determining click confidence degrees of the Y data information based on the second click model and the user interest characteristics, and sequencing the Y data information according to the click confidence degrees of the Y data information;

selecting K pieces of data information from the sequenced Y pieces of data information as electronic recommended reading materials which have an incidence relation with the electronic collection reading materials; and K is an integer less than or equal to Y.

Wherein the method further comprises:

obtaining a positive sample and a negative sample for training an initial click model according to user behavior information associated with each electronic reading in an electronic reading database; the positive sample comprises a click vector pair formed when a click relation exists between a user and the electronic reading; the negative sample comprises an exposure vector pair formed when the user has an exposure relation with the electronic reading material;

training the initial click model based on the positive sample and the negative sample, determining the trained initial click model as a second click model, and adding the second click model to the second recommendation database.

Wherein the target dimension information comprises third dimension information;

pulling target book information associated with the electronic collection reading from a third recommendation database based on the third dimension information;

and taking the pulled target book form information as an electronic recommended reading matter which has an association relation with the electronic collection reading matter.

Wherein the pulling target book information associated with the electronic collection reading from a third recommendation database based on the third dimension information comprises:

searching first book information containing the electronic collection reading materials in a recommendation database based on the third dimension information;

if first book information containing the electronic collection reading materials is searched, the first book information is used as target book information related to the electronic collection reading materials;

if the first book form information containing the electronic collection reading materials is not searched, searching second book form information associated with the subject information of the electronic collection reading materials from the recommendation database, and taking the searched second book form information as target book form information associated with the electronic collection reading materials; the search priority of the first order information is higher than the search priority of the second order information.

Wherein the method further comprises:

performing theme analysis on the theme information of all the electronic readings in the electronic reading database through a theme analysis model to obtain the theme information of each electronic reading;

dividing the clustering cluster to which the subject information of each electronic reading belongs; one cluster corresponds to one topic information;

searching a cluster associated with the subject information of the electronic collection reading materials in the cluster, generating second book order information according to the searched electronic reading materials in the cluster, and adding the second book order information to the third recommendation database;

adding the first book information created based on the operation rule to the third recommendation database.

Wherein the target dimension information includes fourth dimension information;

acquiring the electronic collection reading materials carrying the reading completion identification from the electronic reading material collection column based on the fourth dimension information;

acquiring a completion timestamp of the electronic collection reading materials carrying the read-out identification, sequencing the electronic collection reading materials carrying the read-out identification based on the completion timestamp, and determining the sequenced electronic collection reading materials carrying the read-out identification as the finished electronic reading materials;

and taking the completed electronic reading material and the comment information of the completed electronic reading material as the electronic recommended reading material which has an association relation with the electronic collection reading material.

An aspect of an embodiment of the present application provides an electronic reading data processing apparatus, where the apparatus includes:

the dimension information determining module is used for responding to a first operation triggered aiming at the electronic reading collection column, and determining dimension information corresponding to the first operation as target dimension information in a plurality of dimension information corresponding to the electronic reading collection column; the electronic reading matter collecting column comprises electronic collecting reading matters;

the dimension information sending module is used for sending the target dimension information to a server;

a recommendation list acquisition module, configured to acquire a data push list associated with the target dimension information and returned by the server; the data push list comprises electronic recommended reading materials which are in association relation with the electronic collection reading materials.

An aspect of an embodiment of the present application provides an electronic reading data processing apparatus, where the apparatus is applied to a user terminal, and the apparatus includes: a processor, a memory, and a network interface;

the processor is connected with a memory and a network interface, wherein the network interface is used for providing a data communication function, the memory is used for storing program codes, and the processor is used for calling the program codes to execute the method in one aspect of the embodiment of the application.

An aspect of the embodiments of the present application provides a computer storage medium storing a computer program, where the computer program includes program instructions, and when the processor executes the program instructions, the method according to an aspect of the embodiments of the present application is performed.

the dimension information acquisition module is used for acquiring target dimension information sent by the user terminal; the target dimension information is obtained by a user terminal responding to a first operation aiming at a plurality of dimension information corresponding to the electronic reading collection column; the electronic reading matter collecting column comprises electronic collecting reading matters;

the reading recommending module is used for acquiring the electronic recommended reading which has an association relation with the electronic collection reading based on the target dimension information;

and the recommendation list generation module is used for generating a data push list associated with the target dimension information according to the electronic recommended reading materials and returning the data push list to the user terminal.

Wherein the target dimension information comprises first dimension information;

the reading recommendation module comprises:

the first pulling unit is used for pulling N pieces of data information matched with the electronic collection reading materials from a first recommendation database based on the first dimension information; n is a positive integer;

the first filtering unit is used for filtering S pieces of recommended data information from the N pieces of data information and taking the residual K pieces of data information after filtering as the electronic recommended reading materials which are in incidence relation with the electronic collection reading materials; k ═ N-S and S is an integer greater than or equal to zero.

Wherein the reading recommendation module further comprises:

the system comprises an article vector construction unit, a data processing unit and a data processing unit, wherein the article vector construction unit is used for obtaining an article vector of each electronic reading according to a keyword carried by each electronic reading in an electronic reading database and user information associated with each electronic reading;

the first item determining unit is used for determining the electronic collection reading materials in the electronic reading material database as a first item to obtain an item vector of the first item;

the second item determining unit is used for determining the electronic reading materials except the electronic collection reading materials in the electronic reading material database as second items to obtain item vectors of the second items;

and the similarity determining unit is used for determining cosine similarity between the article vector of the first article and the article vector of the second article, taking the second article with the cosine similarity meeting recommendation conditions as data information matched with the electronic collection reading, and adding the data information matched with the electronic collection reading to the first recommendation database.

Wherein the item vector construction unit includes:

the information acquisition subunit is used for acquiring all the electronic readings in the electronic reading database and user behavior information related to all the electronic readings;

the information analysis subunit is used for analyzing the user behavior information to obtain user information associated with each electronic reading material, and obtaining a user sequence of each electronic reading material according to the user information associated with each electronic reading material;

the word extraction subunit is used for extracting the keywords carried by each electronic reading material and obtaining the keyword sequence of each electronic reading material according to the keywords of each electronic reading material;

and the vector construction subunit is used for constructing the item vector of each electronic reading material based on the user sequence of each electronic reading material and the keyword sequence of each electronic reading material.

Wherein the first pulling unit includes:

a model obtaining subunit, configured to obtain, based on the first dimension information, a first click model associated with the first item;

the candidate determining subunit is used for determining the data information in the first recommendation database into M pieces of candidate data information; m is the total amount of data information in the recommendation database;

a probability value determination subunit configured to determine predicted probability values of the M candidate data information based on the first click model;

the sequencing subunit is used for sequencing the prediction probability values of the M candidate data information, and selecting N data information matched with the electronic collection from the M sequenced candidate data information based on the prediction probability values of the M sequenced candidate data information; n is less than or equal to M.

the reading recommendation module comprises:

the second pulling unit is used for acquiring the classification category to which the electronic collection reading belongs based on the second dimension information and pulling X data information associated with the classification category to which the electronic collection reading belongs from an electronic reading database; x is a positive integer;

the second filtering unit is used for filtering the recommended Z data information from the X data information to obtain the residual Y data information after filtering; y ═ X-Z) and Z is an integer greater than or equal to zero;

and the sequencing unit is used for acquiring the click confidence degrees of the Y data information and determining the electronic recommended reading materials which have an association relation with the electronic collection reading materials from the Y data information according to the click confidence degrees of the Y data information.

Wherein the sorting unit includes:

the confidence determining subunit is used for acquiring a second click model from a second recommendation database, determining the click confidence of the Y data information based on the second click model and the user interest characteristics, and sequencing the Y data information according to the click confidence of the Y data information;

the data selection subunit is used for selecting K pieces of data information from the sorted Y pieces of data information as the electronic recommended reading materials which have an association relationship with the electronic collection reading materials; and K is an integer less than or equal to Y.

Wherein the reading recommendation module further comprises:

the sample determining unit is used for obtaining a positive sample and a negative sample for training an initial click model according to user behavior information associated with each electronic reading in the electronic reading database; the positive sample comprises a click vector pair formed when a click relation exists between a user and the electronic reading; the negative sample comprises an exposure vector pair formed when the user has an exposure relation with the electronic reading material;

and the model training unit is used for training the initial click model based on the positive sample and the negative sample, determining the trained initial click model as a second click model, and adding the second click model to the second recommendation database.

Wherein the target dimension information comprises third dimension information;

the reading recommendation module comprises:

the third pulling unit is used for pulling target book form information related to the electronic collection reading materials from a third recommendation database based on the third dimension information;

and the reading material determining unit is used for taking the pulled target book form information as an electronic recommended reading material which has an association relation with the electronic collection reading material.

Wherein the third pulling unit includes:

the first searching subunit is used for searching first book list information containing the electronic collection reading materials in a recommendation database based on the third dimension information;

the book form determining subunit is used for taking the first book form information as target book form information associated with the electronic collection reading materials if the first book form information containing the electronic collection reading materials is searched;

the book form determining subunit is further configured to search, if first book form information including the electronic collection reading materials is not searched, second book form information associated with the subject information of the electronic collection reading materials from the recommendation database, and use the searched second book form information as target book form information associated with the electronic collection reading materials; the search priority of the first order information is higher than the search priority of the second order information.

Wherein the reading recommendation module further comprises:

the theme analysis unit is used for carrying out theme analysis on the theme information of all the electronic readings in the electronic reading database through a theme analysis model to obtain the theme information of each electronic reading;

the cluster dividing unit is used for dividing the cluster to which the subject information of each electronic reading belongs; one cluster corresponds to one topic information;

the book order generating unit is used for searching the cluster associated with the theme information of the electronic collection reading materials in the cluster, generating second book order information according to the searched electronic reading materials in the cluster, and adding the second book order information to the third recommendation database;

and the book list adding unit is used for adding the first book list information created based on the operation rule to the third recommendation database.

Wherein the target dimension information includes fourth dimension information;

the reading recommendation module comprises:

a reading completion acquisition unit, configured to acquire an electronic collection reading carrying a reading completion identifier from the electronic reading collection column based on the fourth dimension information;

the time stamp sequencing unit is used for acquiring the completion time stamp of the electronic collection reading materials carrying the read-out identification, sequencing the electronic collection reading materials carrying the read-out identification based on the completion time stamp, and determining the sequenced electronic collection reading materials carrying the read-out identification as the finished electronic reading materials;

and the comment determining unit is used for taking the completed electronic reading material and the comment information of the completed electronic reading material as the electronic recommended reading material which has an association relationship with the electronic collection reading material.

An aspect of the embodiments of the present application provides an electronic reading data processing apparatus, where the apparatus is applied to a server, and the apparatus includes: a processor, a memory, and a network interface;

The user terminal in the embodiment of the application can respond to a first operation triggered aiming at an electronic reading matter collection column, and in a plurality of dimension information corresponding to the electronic reading matter collection column, the dimension information corresponding to the first operation is determined as target dimension information; wherein the electronic reading matter collecting column comprises electronic collecting reading matters; the electronic book collection column can be understood as an electronic bookshelf which can be used for storing an electronic book added by a user, and at this time, the electronic book added to the electronic bookshelf can be called as an electronic collection book. Further, the user terminal may send the target dimension information to the server, so that the server may obtain the electronic recommended reading material having an association relationship with the electronic collected reading material based on the target dimension information selected by the user, may further generate a data push list associated with the target dimension information according to the obtained electronic recommended reading material, and returns the data push list to the user terminal. Therefore, through multi-dimensional data recommendation in the application scene of the electronic bookshelf, the interests and hobbies of the user can be fully mined, and then the electronic books fitting the interests and hobbies of the user can be efficiently and accurately matched, so that the book searching efficiency and the book searching accuracy of the user in the bookshelf are improved.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

Fig. 1 is a schematic structural diagram of a network architecture according to an embodiment of the present application;

FIG. 2 is a schematic diagram of a data interaction scenario provided by an embodiment of the present application;

FIG. 3 is a schematic flow chart of a method for processing electronic reading material data according to an embodiment of the present application;

FIG. 4 is a schematic view of an electronic reading material collection column according to an embodiment of the present application;

FIG. 5 is a schematic diagram of an output data push list provided by an embodiment of the present application;

fig. 6 is a schematic diagram of switching dimension information according to an embodiment of the present disclosure;

FIG. 7 is a schematic diagram of another output data push list provided by an embodiment of the present application;

FIG. 8 is a schematic diagram of an embodiment of the present application for viewing an electronic recommended reading under different classification categories;

fig. 9a and 9b are schematic diagrams of still another output data push list provided by an embodiment of the present application;

FIG. 10 is a schematic flow chart of a method for processing electronic reading material data according to an embodiment of the present application;

fig. 11 is a flowchart of a recommendation process for obtaining similar books according to an embodiment of the present application;

FIG. 12 is a schematic flow chart illustrating a process for performing a peer-to-peer goodwill recommendation according to an embodiment of the present disclosure;

FIG. 13 is a schematic flow chart illustrating a recommendation of a book order according to an embodiment of the present disclosure;

FIG. 14 is a schematic flow chart illustrating a read completion recommendation according to an embodiment of the present application;

FIG. 15 is a schematic structural diagram of an electronic reading material data processing device according to an embodiment of the present application;

FIG. 16 is a schematic structural diagram of a computer device according to an embodiment of the present disclosure;

FIG. 17 is a schematic structural diagram of another electronic reading material data processing device provided by the embodiment of the application;

FIG. 18 is a schematic structural diagram of another computer device provided in the embodiments of the present application;

FIG. 19 shows an electronic reading data processing system according to an embodiment of the present application.

Detailed Description

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

Please refer to fig. 1, which is a schematic structural diagram of a network architecture according to an embodiment of the present application. As shown in fig. 1, the network architecture may include a service server 2000 and a user terminal cluster, where the user terminal cluster may include a plurality of user terminals, as shown in fig. 1, and specifically may include a user terminal 3000a, a user terminal 3000b, user terminals 3000c and …, and a user terminal 3000 n; as shown in fig. 1, the user terminal 3000a, the user terminal 3000b, the user terminals 3000c, …, and the user terminal 3000n may be respectively in network connection with the service server 2000, so that each user terminal may perform data interaction with the service server 2000 through the network connection.

As shown in fig. 1, each ue in the ue cluster may be integrally installed with a target application, and when the target application runs in each ue, the target application may perform data interaction with the service server 2000 shown in fig. 1. The target application can be understood as a reading application capable of loading and displaying electronic readings, such as a QQ reading application, a wechat reading application, a nova in a browser, and the like. The electronic reading materials in the embodiment of the application can comprise internet data information such as books, cartoons, magazines and the like; it should be understood that the electronic readings in the target application can include multiple reading modes, for example, the user can select to read the electronic readings in the target application in a text message mode or in an audio message mode.

It should be understood that in order to improve the stay time of the user in the reading application, the accuracy of data recommendation in the reading application needs to be improved, namely, the electronic reading materials according with the interests of the user need to be matched quickly and accurately in the reading application. In view of this, the embodiment of the application considers that multidimensional data recommendation can be performed on electronic readings added to an electronic reading collection column (i.e., a bookshelf) by a user, so as to provide personalized and targeted data recommendation for the user in the reading application, thereby improving the efficiency and accuracy of data recommendation.

For convenience of understanding, in the embodiment of the present application, the target application is a QQ reading application as an example, so as to describe a specific process of implementing user data interaction between the user terminal integrated with the QQ reading application and the service server 2000 through a data display platform corresponding to the QQ reading application. The data display platform can be understood as a business data platform for performing multi-dimensional book recommendation in a bookshelf (i.e., an electronic book collection column) scene, and the data display platform can contain electronic collection books added in the electronic book collection column and also contain a plurality of dimensional information associated with the electronic book collection column, so that different data push lists can be provided for a user in a targeted manner to fit the interests and hobbies of the user. In the embodiment of the application, the electronic reading materials (e.g., books) in the electronic reading material collection column (i.e., the bookshelf) can be referred to as electronic collection reading materials, and the electronic reading materials in the corresponding data push list which is rendered and displayed in the data display platform can be referred to as electronic recommendation reading materials. It should be understood that the electronic recommendation reading material in the embodiment of the present application may be an electronic reading material having an association relationship with an electronic collection reading material.

Further, please refer to fig. 2, which is a schematic view of a data interaction scenario provided in an embodiment of the present application. The user terminal shown in fig. 2 may be the user terminal 3000a in the user terminal cluster shown in fig. 1. Further, the target user as shown in fig. 2 may be understood as a user (e.g., user a) who reads using the QQ reading application in the user terminal 3000 a. As shown in fig. 2, at time M1, a user a may execute a first operation on the electronic reading stock of the QQ reading application, so as to determine, as target dimension information, the dimension information corresponding to the first operation from among a plurality of dimension information corresponding to the electronic reading stock, so as to request a data push list associated with the target dimension information from a service server (i.e., the service server 2000 shown in fig. 1) shown in fig. 2, where the data push list requested by the user terminal may include at least one electronic recommended reading stock having an association relationship with the electronic reading stock.

For convenience of understanding, in the embodiment of the present application, taking the electronic collection in the bookshelf as a book a as an example, when the target user performs a first operation (for example, a click operation) on a plurality of dimension information corresponding to the electronic collection column, the dimension information corresponding to the first operation may be referred to as target dimension information, where the target dimension information may be one or more of the plurality of dimension information, and this is not limited here. For convenience of understanding, in the embodiment of the present application, for example, the target dimension information is one of the plurality of dimension information, as shown in fig. 2, the service server may receive the target dimension information sent by the user terminal shown in fig. 2, that is, the service server may receive a data pull request carrying the target dimension information, and may further obtain, based on the target dimension information, the electronic recommended reading having an association relationship with the electronic collection from the recommendation database shown in fig. 2, and may further generate a data push list according to the obtained electronic recommended reading. As shown in fig. 2, the service server may return the generated data inference list to the user terminal shown in fig. 2, so as to output and display the electronic recommended reading materials having an association relationship with the book a in the user terminal shown in fig. 2. In the embodiment of the application, an interface where the target dimension information is located may be referred to as a data push interface, and the electronic recommended reading 10a, the electronic recommended reading 10b, and the electronic recommended reading 10c shown in fig. 2 may be specifically displayed in the data push interface.

The multiple pieces of dimension information shown in fig. 2 may include first dimension information, second dimension information, third dimension information, and fourth dimension information. For each dimension of information on the data presentation platform, a different data push list may be requested. For example, taking the electronic collection in the electronic book collection column (i.e. bookshelf) as book a, the data push list associated with the requested target dimension information may include: related books similar to the content and user group of the book a, books similar to the classification category of the book a, book form information similar to the subject of the book a, comment information when the identification of the book a is read out, and the like.

For example, when the target user performs a refreshing operation (or a clicking operation) on the first dimension information (e.g., the first dimension information in the data pushing interface) on the data display platform, the service server may quickly recommend to the target user a related book, such as book B, book C, and the like, which has similar book content to the book a, at this time, a list in which the related book, such as book B, book C, and the like, is located may be referred to as a data pushing list associated with the first dimension information, and the related books, such as book B and data C, may be referred to as electronic recommended books having an association relationship with the book a.

For example, when the target user performs a click operation on the second dimension information on the data display platform, the service server may recommend books such as book B, book D, etc. having the same classification category as book a to the target user, at this time, a list in which books such as book B, book D, etc. having the same classification category are located may be referred to as a data push list associated with the second dimension information, and may collectively refer to books such as book B and data D having the same classification category as an electronic recommended reading matter having an association relationship with book a.

The third dimension information in this embodiment may be used to recommend target book list information associated with a book (e.g., book a) in the bookshelf to the target user, for example, the business server may recommend book list information (e.g., book list a) containing the book a to the target user. Optionally, the business server may also recommend the target user with book information (e.g., book B) similar to the subject information (referred to as subject) of the book a. In the embodiment of the application, the book sheet a containing the book a or the book sheet B with a theme similar to that of the book a may be referred to as target book sheet information, and a plurality of books in the target book sheet information may be referred to as electronic recommended books having an association relationship with the book a. It should be understood that, in the embodiment of the present application, the book information including the book a may be referred to as first book information, and the first book information may be a book list constructed based on the operation rule. In addition, the book form information similar to the theme of the book a may be referred to as second book form information, where the second book form information is a book form list constructed by performing theme analysis through a theme analysis algorithm. The first order information and the second order information may be stored in the recommendation database, and the search priority of the first order information may be higher than the search priority of the second order information.

The fourth dimension information in the embodiment of the application may be used to recommend a read-out book list to the target user, where the read-out book list may include the electronic collected reading that the target user has read out and comment information of the read-out electronic collected reading. The electronic collected reading materials which are completely read in the electronic reading material collecting column by the target user can be collectively called electronic collected reading materials with completely read marks, the electronic collected reading materials with completely read marks can be collectively called completed electronic reading materials, and at the moment, the business server can refer the completed electronic reading materials and the comment information of the completed electronic reading materials as electronic recommended reading materials which have an association relationship with the book A with completely read marks.

It can be understood that when the target dimension information associated with the electronic reading material favorite column is any one of the plurality of dimension information shown in fig. 2, the target user can freely switch the dimensions among the plurality of dimensions shown in fig. 2, so that data recommendation results of different dimension information can be provided for the target user in the user terminal. For example, the target user may switch the current dimension information from the first dimension information to the second dimension information, so that the electronic recommended reading materials in the data push list associated with the second dimension information may be displayed on the data presentation platform. Of course, the target user may also select other dimension information on the data presentation platform, so that the electronic recommended readings in the data push list associated with the other dimension information may be output, which is not listed here.

The specific process of the service server obtaining the data push list associated with the target dimension information may refer to the following embodiments corresponding to fig. 3 to fig. 14.

Further, please refer to fig. 3, which is a flowchart illustrating a method for processing electronic reading material data according to an embodiment of the present application. As shown in fig. 3, the terminal involved in the method may include a user terminal and a server, and the method may be applied to perform multidimensional data recommendation in an electronic bookshelf scene, and the method at least includes:

step S101, a user terminal responds to a first operation triggered aiming at an electronic reading collection column, and in a plurality of dimension information corresponding to the electronic reading collection column, the dimension information corresponding to the first operation is determined as target dimension information;

specifically, when starting a target application (e.g., a reading application), the user terminal may receive, in the reading application, a first operation triggered by a target user for an electronic reading collection column, and refer dimension information corresponding to the first operation to be target dimension information among a plurality of dimension information corresponding to the electronic reading collection column.

The electronic reading matter collecting column can be understood as a container for storing electronic reading matters in the reading application, in this case, the container for storing electronic reading matters can be referred to as an electronic bookshelf, the electronic bookshelf can be referred to as a bookshelf in this application embodiment, and the electronic reading matters added and stored in the bookshelf can be referred to as electronic collecting reading matters. In addition, the display interface to which the electronic reading collection column belongs can be called a collection interface in the embodiment of the application, and the collection interface can be one interface in the data display platform. After the target user accesses the reading application, the target user can continue to read the electronic collection reading added in the electronic reading collection column in the collection interface. Optionally, after the target user accesses the reading-class application, multi-dimensional data recommendation can be performed by triggering the target area in the collection interface.

It should be understood that the plurality of dimension information corresponding to the electronic reading material collection column may include the first dimension information, the second dimension information, the third dimension information, and the fourth dimension information. Therefore, after the target user accesses the reading application, the user terminal can output a collection interface to which the electronic reading collection bar in the reading application belongs. At this time, the target user can continue to read the collected electronic reading materials in the collection interface (the collected electronic reading materials can be called as electronic collection reading materials), so that the electronic collection reading materials which are attached with the interests of the target user can be quickly acquired in the electronic reading material collection column, and the target user can quickly jump to the reading interface where the corresponding progress is located to continue reading based on the reading progress of the electronic collection reading materials.

For easy understanding, please refer to fig. 4, which is a schematic diagram of an electronic reading matter collecting column according to an embodiment of the present application. When the target user triggers a reading application (for example, a QQ reading application) in the user terminal, the interface 100a shown in fig. 4 may be displayed in the user terminal, and the interface 100a may be a collection interface to which an electronic reading material collection bar shown in fig. 4 (i.e., a bookshelf shown in fig. 4) belongs. As shown in fig. 4, the favorite interface (i.e., the interface 100a) may include a plurality of electronic favorite readers, which may specifically include: fig. 4 shows an electronic collection of

readings

20a, 20b, 20c, 20 d. It is understood that the electronic

collection reading materials

20a, 20b, 20c and 20d shown in fig. 4 may be understood as part of the plurality of electronic collection reading materials added to the bookshelf, and the target user may perform a sliding operation (e.g., sliding down) in the interface 100a shown in fig. 4 to obtain more electronic collection reading materials.

As shown in fig. 4, the target user may select any one of the electronic collections for reading continuously in the page 100a shown in fig. 4, for example, may select to read continuously the electronic collection 20a in the page 100a shown in fig. 4, where the electronic collection 20a may be a book with a book name AAAAA. It is understood that the interface 100a shown in fig. 4 may also display the reading progress of each electronic collection. In other words, the user terminal may record the reading progress of each electronic collection reading in the electronic reading collection column, and may perform progress management on each electronic collection reading according to the reading progress of each electronic reading (for example, the reading progress of a book with a book name of AAAAA may be 10% of what has been read), so that a friendly continuous reading manner may be provided for the target user through the recorded reading progress, so as to enhance the user viscosity.

It can be understood that, when obtaining the reading progress of each electronic collection, the user terminal may further set, as a reading completion flag, the flag of the electronic collection (e.g., the electronic collection 20c shown in fig. 4) whose reading progress reaches the progress threshold (e.g., 100%) in the electronic collections, and may record the completion timestamp of the electronic collection carrying the reading completion flag, so that a data push list associated with the electronic collection carrying the reading completion flag may be subsequently constructed in the server.

Optionally, as shown in fig. 4, after the target user accesses the reading-class application, a click operation may be performed on the target area in the favorite interface (i.e., the interface 100a shown in fig. 4), and at this time, the click operation triggered for the favorite interface may be referred to as a first operation in the embodiment of the present application. Meanwhile, the user terminal may respond to a first operation triggered with respect to the target area in the electronic reading collection bar, and refer to preset dimension information (e.g., first dimension information) bound for the target area in advance as dimension information corresponding to the first operation, where the dimension information (i.e., the first dimension information) corresponding to the first operation may be collectively referred to as target dimension information, so that step S102 may be further performed subsequently.

Step S102, the user terminal sends the target dimension information to a server.

It is understood that, after the user terminal performs the step S101, the user terminal may send target dimension information (for example, the default dimension information bound to the target area is the first dimension information) associated with the target area to the server, so that the server may further perform the step S103. The server may be the service server 2000 in the embodiment corresponding to fig. 1.

Step S103, the server acquires electronic recommended reading materials which are in incidence relation with the electronic collection reading materials based on the target dimension information;

the server may have an online recommendation function, that is, when a data pull request sent by a user terminal is received, electronic recommended books having an association relationship with books in the bookshelf may be pushed online to a target user according to target dimension information.

Optionally, the server may further have an offline processing function, for example, taking an example that the plurality of pieces of dimension information corresponding to the electronic reading collection bar include first dimension information, second dimension information, third dimension information, and fourth dimension information, the server may pre-process to obtain a first recommendation database associated with the first dimension information, a second recommendation database associated with the second dimension information, a third recommendation database associated with the third dimension information, and a fourth recommendation database associated with the fourth dimension information. It can be understood that the first recommendation database, the second recommendation database, the third recommendation database, and the fourth recommendation database are recommendation databases for a target user, and the recommendation database for the target user may be the recommendation database in the embodiment corresponding to fig. 2.

Wherein the first recommendation database may be used to store other books similar to the book added to the bookshelf by the target user. The second recommendation database may be used to store books similar to the classification category of the book that the target user added to the bookshelf. The third recommendation database may be configured to store the first book order information and the second book order information, where the first book order information may be understood as an operation book order constructed based on operation rules, and the second book order information may be understood as a book order similar to a theme of a book added to the bookshelf by the target user. The fourth recommendation database may be configured to store comment information of books with read identifiers in the bookshelves.

For ease of understanding, please further refer to the recommendation database shown in FIG. 4 above. The recommendation database shown in fig. 4 may include a plurality of recommendation databases, and the plurality of recommendation databases may specifically include the recommendation database 10, the recommendation database 20, the recommendation database 30, and the recommendation database 40 shown in fig. 4. The recommendation database 10 may be the first recommendation database, the recommendation database 20 may be the second recommendation database, the recommendation database 30 may be the third recommendation database, and the recommendation database 40 may be the fourth recommendation database. The electronic readings in the recommendation database shown in fig. 4 can be collectively referred to as electronic recommendation readings in the embodiment of the present application.

It should be understood that the electronic recommended reading materials in the plurality of recommended databases are determined by the server in advance according to the recommendation rules corresponding to the different dimension information in the user terminal, that is, the server may perform offline analysis on the books in the bookshelf of the target user in advance according to the corresponding recommendation rules to obtain the recommended databases associated with the different dimension information. The electronic reading database can be used for storing all electronic readings associated with the reading application (for example, the QQ reading application), such as Internet data information of electronic novels, electronic books, electronic magazines and the like.

104, the server generates a data push list associated with the target dimension information according to the electronic recommended reading materials;

specifically, after the server in this embodiment of the application finishes executing the step S103, the template information associated with the target dimension information may be further acquired, for example, the template information associated with the target dimension information may include any one of the foregoing template information associated with the first dimension information, template information associated with the second dimension information, template information associated with the third dimension information, and template information associated with the fourth dimension information, so that the electronic recommended reading obtained in the step S103 may be integrated according to the template information associated with the target dimension information, so as to obtain the data push list associated with the target dimension information. Further, the server may execute steps S105 to S106, so that the user terminal can quickly output the electronic recommended reading materials having an association relationship with the electronic favorite reading materials when obtaining the corresponding data push list.

And step S105, the server returns the data push list to the user terminal.

And S106, outputting the electronic recommended reading materials which are contained in the data list and have the association relation with the electronic collection reading materials by the user terminal.

It should be understood that, in this embodiment of the application, if the target dimension information includes the first dimension information, the server may perform intelligent layout on the electronic recommended reading materials (e.g., N selected similar books, where N is an integer greater than zero) related to the electronic collected reading materials (i.e., books in the bookshelf) according to the template information associated with the first dimension information to form a data push list associated with the first dimension information, so that step S105 may be further performed, so that the user terminal may quickly render and display the data push list associated with the first dimension information in the data push interface to which the first dimension information (e.g., the recommended items bound to the target area are book items) belongs. By analogy, the embodiment of the application can also quickly render and display the data push list associated with the second dimension information in the data push interface to which the second dimension information (for example, the recommended item bound with the target area is a similar looking item) belongs. By analogy, the embodiment of the application can also quickly render and display the data push list associated with the third dimension information in the data push interface to which the third dimension information (for example, the recommended item bound with the target area is a book entry). By analogy, the embodiment of the application can also quickly render and display the data push list associated with the fourth dimension information in the data push interface to which the fourth dimension information (for example, the recommended entry bound with the target area is a read-out entry) belongs.

For convenience of understanding, in the embodiment of the present application, the interface 100a in the embodiment corresponding to fig. 4 is still taken as an example to illustrate a specific process of outputting, in the user terminal, the data push list associated with different dimension information after the target user clicks the target area in the embodiment corresponding to fig. 4, so that multi-dimensional data recommendation can be quickly and accurately implemented in a bookshelf scene.

Further, please refer to fig. 5, which is a schematic diagram of an output data push list according to an embodiment of the present application. The interface 100a shown in fig. 5 may be the interface 100a in the embodiment corresponding to fig. 4, and in this case, the electronic collection reading in the electronic reading collection column may include the electronic collection reading 20a, the electronic collection reading 20b, the electronic collection reading 20c, and the electronic collection reading 20d shown in fig. 5. After the target user clicks the target area in the electronic reading material collection column, the interface 200a shown in fig. 5 may be skipped, where the interface 200a may be a data push interface shown in fig. 5 and associated with the electronic reading material collection column, that is, a plurality of recommendation entries may be displayed in the interface 200a shown in fig. 5, where the recommendation entries may specifically include "book" entry, "good-looking-at-the-same" entry, "book" entry, "read-completed" entry shown in fig. 5. It should be understood that, the plurality of recommendation entries displayed in the interface 200a are the plurality of dimension information corresponding to the electronic reading material collection bar, and one recommendation entry may be equivalent to one dimension information. Therefore, the dimension information bound with the target area shown in fig. 5 may be the first dimension information shown in fig. 5, that is, the recommended item bound with the target area may be the "book" item shown in fig. 5. It should be understood that based on the data interaction method described in the foregoing steps S101-S106, the data push list under the book entry can be quickly output and displayed in the interface 200a shown in fig. 5. The data push list under the "book" entry may specifically include the electronic recommended reading 30a, the electronic recommended reading 30b, the electronic recommended reading 30c, and the electronic recommended reading 30d shown in fig. 5. For example, the electronic recommended reading materials having an association relationship with the electronic collection reading material 20a (i.e., the book with the book name AAAAA in the bookshelf) may include the electronic recommended reading material 30a and the electronic recommended reading material 30b shown in fig. 5. At this time, the electronic recommended reading 30a, 30b displayed in the page 200a may have similar contents to the electronic collection reading 20a in the bookshelf. For another example, the electronic recommended reading materials associated with the electronic collection reading materials 20b (i.e., the books with the book name AABBA in the bookshelf) may include the electronic recommended reading materials 30c and the electronic recommended reading materials 30d shown in fig. 5, which are not listed here.

It should be understood that, as shown in fig. 5, since the template information associated with the first dimension information may be in a portrait mode, the electronic recommended reading currently displayed in the interface 200a may contain a related book that is partially similar to the book in the bookshelf, and in view of this, the target user may further view more similar books by clicking a function button (i.e., a pull-down button at the lower left corner in the interface 200 a) in the interface 200a shown in fig. 5.

It should be appreciated that the target user may dimension switch between multiple recommendation entries (i.e., multiple dimension information) as shown in FIG. 5 to quickly switch and display data push lists associated with different dimension information.

For easy understanding, please refer to fig. 6, which is a schematic diagram of switching dimension information provided in an embodiment of the present application. In the embodiment of the present application, the interface 200a shown in fig. 6 is obtained by taking the data push list associated with the first dimension information in the interface 200a shown in fig. 5 as an example. As shown in fig. 6, the target user may trigger the "good-looking-at-the-same-class" entry (i.e., the second dimension information) in the interface 200a shown in fig. 6, so that the user terminal may respond to the new first operation triggered by the "good-looking-at-the-same-class" entry, and use the "good-looking-at-the-same-class" entry as the new target dimension information, so as to quickly obtain the data push list associated with the second dimension information from the server according to the above steps S101 to S105, and output and display the data push list associated with the second dimension information in the interface 300a shown in fig. 6.

It should be understood that the interface 200a and the interface 300a shown in fig. 6 are data pushing interfaces under different dimensional information, respectively, so that in the embodiment of the present application, the data pushing interfaces under different dimensional information may be collectively referred to as data pushing interfaces associated with a bookshelf, and the data pushing interface may also be an interface in a data presentation platform of the reading application. Therefore, when a click operation executed on a data push interface associated with an electronic reading material favorite bar is received, the click operation is referred to as a new first operation or as a second operation, so that dimension information corresponding to the new first operation can be determined as new target dimension information (for example, the new target dimension information may be the second dimension information shown in fig. 6), and a data push list associated with the new target dimension information can be quickly output. When the server receives the data pulling request, the server can obtain the corresponding data push list in a targeted manner, so that the user terminal can be effectively prevented from pulling too much recommended data at one time, and the waste of data flow in the user terminal can be avoided.

As shown in fig. 6, the data push list associated with the second dimension information is determined by the classification category to which the electronic book collection belongs, that is, when the server divides the classification categories of the electronic book collection in the bookshelf, a plurality of sub-data push lists may be formed based on the number of the divided classification categories (that is, 4 classification categories shown in fig. 6), so that it may be ensured that one classification category may correspond to one sub-data push list. It should be understood that a sub data push list can be understood as a sort card containing the same sort category, i.e., electronic push books in different sub data push lists (i.e., different sort cards) have different sort categories. The sub-data push lists corresponding to the classification categories displayed in the interface 300a shown in fig. 6 may be collectively referred to as a data push list associated with the second dimension information. Similarly, the target user may perform a sliding operation on each displayed category card in the aforementioned interface 300a shown in fig. 6 to view the electronic recommended readings in different category cards through a sliding operation manner (for example, a sliding manner up and down), so as to achieve quick review of the recommendation data of different category.

Optionally, if the dimension information bound to the target area in the interface 100a is second dimension information, when the target user performs a first operation on the target area in the electronic reading collection bar, the second dimension information corresponding to the first operation may be referred to as target dimension information, so that the data push list associated with the second dimension information acquired from the server may be quickly rendered and displayed in the user terminal based on the steps S101 to S106.

For easy understanding, please refer to fig. 7, which is a schematic diagram of another output data push list provided in the embodiment of the present application. As shown in fig. 7, the target user may perform a first operation on the target area in the interface 100b shown in fig. 7, and at this time, the dimension information corresponding to the first operation may be second dimension information bound to the target area in the interface 100b in advance. In view of this, the user terminal in the embodiment of the present application may quickly acquire the data push list associated with the second dimension from the server based on the foregoing steps S101 to S106, and display the data push list associated with the second dimension information in the display interface 200b shown in fig. 7. At this time, the data push list associated with the second dimension information may specifically include sub data push lists corresponding to a plurality of classification categories shown in fig. 7. As shown in fig. 7, the sub-data push list corresponding to the classification category 4 may specifically include the electronic recommended reading 40a, the recommended electronic reading 40b, and the recommended electronic reading 40c shown in fig. 7. It can be understood that, in the embodiment of the present application, the sub-data push list corresponding to the classification category 4 may be referred to as a classification card 1, and each electronic recommended reading in the classification card 1 may be separately displayed according to a display mode of a marketing card, specifically, refer to the layout diagram of the electronic recommended reading under the classification category 4 shown in fig. 7.

It is to be understood that if Y (for example, 6) books having the same classification category as the electronic collection (for example, the electronic collection 20a) can be displayed in the template information corresponding to the second dimension information, the electronic recommended reading associated with the corresponding classification category may be displayed by sliding up and down in the interface 200b shown in fig. 7.

For easy understanding, please refer to fig. 8, which is a schematic diagram illustrating an example of the present application for viewing the electronic recommended reading materials under different classification categories. For the 4 classification categories shown in fig. 8, the target user can view the electronic recommended reading materials under different classifications by sliding up and down. As shown in fig. 8, the target user may slide the classification card 1 to which the classification category 4 belongs upward, and Y pieces of electronic recommended readings in the classification card 1 may be displayed in the user terminal. At this time, another Y (i.e. another 6) electronic recommended readings can be displayed by clicking the "change over" button in the classification card, that is, the recommended electronic readings can be filtered through the bloom filter, so that repeated recommendation of data can be avoided.

As shown in fig. 8, optionally, the target user may also hide the electronic recommended reading 40a, the recommended electronic reading 40b, and the recommended electronic reading 40c in the classification category 4 by sliding the classification card 1 to which the classification category 4 belongs downward. Meanwhile, the user terminal can also synchronously display the sub-data push list corresponding to the classification type 3, namely the sub-data push list corresponding to the classification type 3 can be called a classification card 2, and Y electronic recommended readings in the classification card 2 can be displayed. It is understood that the classification category of the Y electronic recommended readings in the classification card 2 shown in fig. 8 is different from the classification category of the Y electronic recommended readings in the classification card 1. By analogy, the sub-data push list corresponding to the classification category 2 may be referred to as a classification card 3, and the sub-data push list corresponding to the classification category 1 may be referred to as a classification card 4. In view of this, the embodiment of the application may also display the Y electronic recommended readings in the classification card 3 to which the classification category 3 belongs, in synchronization when the Y electronic recommended readings in the classification card 2 are hidden. By analogy, when the Y electronic recommended readings in the classification card 3 are hidden, the Y electronic recommended readings in the classification card 4 belonging to the classification category 1 can be synchronously displayed.

Optionally, when the dimension information bound to the target area in the page 100a is the third dimension information, the data push list associated with the third dimension information may be quickly acquired from the server, and the data push list associated with the third dimension information may be displayed in the user terminal. Similarly, when the dimension information bound to the target area in the page 100a is the fourth dimension information, the data push list associated with the fourth dimension information may be quickly obtained from the server, and the data push list associated with the fourth dimension information may be displayed in the user terminal.

For easy understanding, please refer to fig. 9a and fig. 9b, which are schematic diagrams of another output data push list provided in an embodiment of the present application. The dimension information bound to the target area in the interface 100c shown in fig. 9a may be third dimension information, and when the target user performs the first operation on the target area in the electronic reading collection box, based on the foregoing steps S101 to S105, the data push list associated with the third dimension information may be quickly acquired from the server, and the data push list associated with the third dimension information may be displayed in the interface 200c shown in fig. 9 a. The data push list associated with the third dimension information may include book list information where books in the bookshelf are located (for example, the book list information where the electronic collection 20a is located may be the first book list information). The first book information may include two book information shown in fig. 9a, the two book information may be an operation book constructed by an operator in advance according to an operation rule, and both the two operation books may include the electronic collection 20 a. At this time, the embodiment of the present application may refer to the two pieces of book form information as target book form information. Optionally, if the recommendation database associated with the third dimension information (i.e., the third recommendation database) does not have the first book list information containing the books in the bookshelf, it is necessary to further search for second book list information having a similar theme with the books in the bookshelf, and refer to the second book list information having a similar theme with the books in the bookshelf as the target book list information, so as to push the book list information having similar theme information to the target user, thereby providing more novelty and surprise for the user.

The dimension information bound to the target area in the interface 100d shown in fig. 9b may be fourth dimension information, and when the target user performs the first operation on the target area in the electronic reading collection box, based on the foregoing steps S101 to S105, the data push list associated with the fourth dimension information may be quickly acquired from the server, and the data push list associated with the fourth dimension information may be displayed in the interface 200d shown in fig. 9 b. Wherein, the data push list associated with the fourth dimension information may include books in the bookshelves that have been read by the target user. As shown in fig. 9b, the electronic collection read by the target user may include the electronic collection 20c shown in fig. 9b, and may further include an electronic collection 20d not listed in the interface 100d, as shown in fig. 9b, the book name of the electronic collection 20c may be "AABBCC", and the target user has already published a comment on the electronic collection 20c, so that the comment information 1 shown in fig. 9b may be pulled from the fourth recommendation database associated with the fourth dimension information. In addition, for the case that the title of the electronic collection 20d may be "CCDCC", the target user may not have published a comment for the electronic collection 20d, so that the comment information 2 shown in fig. 9b may be pulled from the fourth recommendation database associated with the fourth dimension information, and the comment information 2 may be the first two pieces of comment information with the highest scores that are previously screened out by the server from all pieces of comment information for the electronic collection 20d (for example, the higher the number of praise is, the higher the score of the comment information is, so that the comment information with the higher score may be referred to as good comment information).

Therefore, by combining the application scene of the bookshelf, a plurality of data information with different dimensions can be recommended to the target user quickly and accurately, and the diversity of data recommendation can be enriched. In addition, each recommendation database is obtained offline, and the data recommendation efficiency can be improved in the online recommendation process.

For easy understanding, please refer to fig. 10, which is a flowchart illustrating an electronic reading data processing method according to an embodiment of the present application. The method can be applied to a server, and comprises the following steps:

step S201, acquiring target dimension information sent by a user terminal;

the target dimension information is obtained by a user terminal responding to a first operation aiming at a plurality of dimension information corresponding to the electronic reading matter collecting column; the electronic reading matter collecting column comprises electronic collecting reading matters. The dimension information may include the first dimension information, the second dimension information, the third dimension information, and the fourth dimension information.

For a specific process of the server acquiring the target dimension information sent by the user terminal, reference may be made to the description of step S101 to step S102 in the embodiment corresponding to fig. 1, which will not be further described here.

Step S202, acquiring electronic recommended reading materials which are in incidence relation with the electronic collection reading materials based on the target dimension information;

the electronic collection reading materials can contain internet data information such as books, magazines, cartoons and the like. For the convenience of understanding, in the embodiment of the present application, taking an electronic collection as an example of a book, the first dimension information may be used to recommend, to the target user, a book related to the book in the bookshelves, for example, taking the book in the bookshelves as book a, and recommend, to the target user, a related book similar to the content of the book a (for example, books with similar content, such as book a1, book a2, and the like). The second dimension information may be used to recommend books to the target user, which are the same as the classification categories of the books in the bookshelves, for example, taking the books in the bookshelves as book B, the classification categories of the book B may be the category labels of the story novels, so that the books carrying the category labels of the story novels (e.g., books with the same classification categories such as book B1, book B2, etc.) may be recommended to the target user. The third dimension information may be used to recommend books similar to the subject of the books in the bookshelf to the target user, for example, taking the books in the bookshelf as book C, the subject of the book C may be the overlord overall theme information, so that the order information with the overlord overall theme information may be recommended to the target user (for example, books such as book C1 and book C2 in the order information all have the same subject information as book C in the bookshelf). The fourth dimension information can be used for recommending comment information of books with read marks in the bookshelf to the target user.

And step S203, generating a data push list associated with the target dimension information according to the electronic recommended reading materials, and returning the data push list to the user terminal.

When the target dimension information includes the first dimension information, a specific process of the server acquiring the electronic recommended reading material having an association relationship with the electronic collection reading material based on the target dimension information may be described as follows: the server can pull N pieces of data information matched with the electronic collection reading materials from a first recommendation database based on the first dimension information; n is a positive integer; further, the server may filter S recommended data information from the N data information, and use K remaining data information after filtering as the electronic recommended reading material having an association relationship with the electronic collection reading material, so as to further perform step S203; wherein K is (N-S) and S is an integer greater than or equal to zero.

Before recommending the N books with the highest similarity to the books in the bookshelf to the target user, data information (namely offline big data obtained by offline processing) matched with the books in the bookshelf needs to be added to the first recommendation database in advance for storage, so that the server can quickly pull the N data information matched with the electronic collected reading materials from the M candidate data information contained in the first recommendation database when receiving a data pull request initiated by the user terminal, wherein N may be less than or equal to M.

In other words, the server can obtain the item vector of each electronic reading material in advance according to the keywords carried by each electronic reading material in the electronic reading material database and the user information associated with each electronic reading material; further, the server determines the electronic collection reading materials in the electronic reading material database as a first article to obtain an article vector of the first article; further, the server determines the electronic reading materials except the electronic collection reading materials in the electronic reading material database as a second article to obtain an article vector of the second article; further, the server determines cosine similarity between the article vector of the first article and the article vector of the second article, takes the second article with the cosine similarity meeting recommendation conditions as data information matched with the electronic collection reading materials, and adds the data information matched with the electronic collection reading materials to the first recommendation database, so that when a subsequent server receives a data pulling request, the data information matched with the electronic collection reading materials can be collectively referred to as M candidate data information in the first recommendation database, and M is the total number of the data information matched with the electronic collection reading materials in the bookshelf.

For convenience of understanding, in the embodiment of the present application, the electronic

collection reading materials

20a, 20b, 20c, and 20d in the electronic reading material collection column (i.e., the bookshelf) in the embodiment corresponding to fig. 5 are taken as examples to illustrate a specific process of acquiring similar books having an association relationship with books in the bookshelf. Further, please refer to fig. 11, which is a flowchart illustrating a recommendation process for obtaining similar books according to an embodiment of the present application. As shown in fig. 11, the target user may execute step S11 in the user terminal, that is, may execute a click operation in the page 100a (i.e., the favorite page) in the embodiment corresponding to fig. 5 or execute a refresh operation in the interface 200a (i.e., the data push page) in the embodiment corresponding to fig. 5, so as to send a data pull request to the server shown in fig. 11. At this time, the server may perform step S14, that is, N (for example, N1+ N2+ N3+ N4 is 12) electronic readings similar to the book in the bookshelf may be pulled in real time from M (for example, 100) candidate data information included in the first recommended database described in step S15 shown in fig. 11 based on the first dimension information carried in the data pull request. The N electronic readings in the embodiment of the application are N data information with higher prediction probability values, which are screened out after the server carries out online probability prediction and sequencing on the M candidate data information based on the first click model. The first click model may be a click rate ranking model trained by the server based on user click behaviors (i.e., click relationships between users and books, etc.). For example, the server may pull n1 (e.g., 3) data messages having a height similar to the electronic collection 20a, n2 (e.g., 3) data messages having a height similar to the electronic collection 20b, n3 (e.g., 3) data messages having a height similar to the electronic collection 20c, and n4 (e.g., 3) data messages having a height similar to the electronic collection 20 d. Wherein n1, n2, n3 and n4 may be equal or unequal.

Further, as shown in fig. 11, the server may further perform step S13, that is, the electronic recommended reading that is issued by the click operation or the pull-down refresh operation and displayed in the user terminal may be filtered by a bloom filter, so as to prevent the electronic reading data from being recommended repeatedly. In other words, the embodiment of the application can update the pull record and the exposure record of the target user for the books (namely the electronic collection books) in the bookshelf to the bloom filter in real time after the user terminal pulls and displays the electronic recommended books each time, so as to prevent repeated recommendation of electronic book data. For example, S pieces of data information (for example, 2 pieces of data, where the 2 pieces of data information may be data information displayed in the interface 200a shown in fig. 5 for the last second exposure) that has been recommended may be filtered from the aforementioned N pieces of data information (for example, 2 pieces of data), so that K pieces of data information (i.e., 10 pieces of data information) remaining after filtering may be taken as the electronic recommended reading having an association relationship with the electronic favorite, and thus a book push list associated with a book in a bookshelf may be generated based on the K pieces of data information, so that step S12 may be executed in the user terminal to pull and display the K pieces of similar books in the data push list.

In practical engineering, in order to avoid that the electronic recommended readings currently displayed in the interface 200a in the embodiment corresponding to fig. 5 are all K books similar to the electronic collected readings 20a, books similar to a plurality of books in the bookshelf may be presented in the first recommended database in a classified and scattered manner as much as possible, for example, ni books similar to the electronic collected readings in the bookshelf may be pre-screened in the first recommended database, where the value of i is determined by the classification to which the electronic collected readings in the bookshelf belong. For example, if the classification categories of the four electronic collections are different, the book with the front predicted probability value of 10 (i.e., N1) books can be extracted from the books with the front predicted probability value of M1 books (e.g., mi is 20 books, and these 20 books can be partial data information in the M book candidate data information) similar to the electronic collection 20c in the first recommendation database, and so on, the ni books with each electronic collection similar to the electronic collection in the bookshelf can be obtained, and the N data information matching the electronic collection can be quickly extracted.

Before executing the foregoing step S15, as shown in fig. 11, the server further needs to calculate the similarity of the contents of the two articles based on the collaborative filtering model of the articles, and then recommend M articles similar to each other according to the articles that the user joins the bookshelf. The article-based collaborative filtering model may first calculate a similarity between two articles, and then recommend similar articles for the articles according to the articles added to the bookshelf by the user (for example, may recommend M articles with similar contents for the articles). In other words, the server may perform steps S20-S15 shown in fig. 11 in advance to store M books similar to the book in the bookshelf of the target user offline into the first recommendation database. The M similar books added into the first recommendation database are all data information of which the cosine similarity meets the recommendation condition (for example, the cosine similarity with the books in the bookshelf is greater than a preset cosine threshold value). As shown in fig. 11, the server may periodically (for example, periodically in units of days) acquire the user behavior flow reported by each user, and analyze the user behavior flow to obtain the user information associated with each electronic reading in the electronic reading database, that is, the user information associated with each electronic reading in each book on the same day may be obtained, for example, the user information associated with the electronic reading may be understood that n (for example, 100) people click to read the electronic collection 20c today counted by the server offline. Of course, the server can also count the user information associated with other electronic readings in the electronic readings database.

As shown in fig. 11, after the server completes step S18, the server may obtain the user information associated with each electronic reading material to obtain the user sequence of each electronic reading material. In addition, m keywords carried by each electronic reading can be obtained through the foregoing step S19, so as to obtain a keyword sequence of each electronic reading. Further, step 17 shown in fig. 11 indicates that the server can construct the item vector of each electronic reading according to the user sequence of each electronic reading and the keyword sequence of each electronic reading. Wherein, the item vector of the electronic reading is equal to (user 1, user 2, …, user n, word 1, word 2. It should be understood that the value of each dimension in the item vector may be 0 or 1, where a value of a user dimension of 1 may indicate that the user has a reading behavior on the electronic reading material, and a value of 0 indicates that the user does not have reading information on the electronic reading material. Similarly, the value of the word dimension may also be 0 or 1, where a value of 1 indicates that a keyword is extracted, and a value of 0 indicates that a keyword is not extracted.

Therefore, the server can obtain the item vector of each electronic reading material in advance based on the keywords carried by each electronic reading material and the user information associated with each electronic reading material. Therefore, M books similar to the book in the bookshelf in any one of the user terminals can be determined by the calculation formula of the cosine similarity in step S16 shown in fig. 11. For example, for the user terminal shown in fig. 11, M pieces of candidate data information matching books (i.e., the electronic collected material 20a, the electronic collected material 20b, the electronic collected material 20c, and the electronic collected material 20d) in a bookshelf in the user terminal can be obtained in advance through the collaborative filtering model.

In the embodiment of the present application, when the item vector of each electronic reading is obtained, the electronic reading (the electronic reading identical to the book in the bookshelf) identical to the electronic collection reading 20a, the electronic collection reading 20b, the electronic collection reading 20c, and the electronic collection reading 20d in the electronic reading database is referred to as a first item, and the electronic reading other than the electronic collection reading 20a, the electronic collection reading 20b, the electronic collection reading 20c, and the electronic collection reading 20d in the electronic reading database is referred to as a second item, so as to obtain the cosine similarity between the two items based on the item vector of the first item and the item vector of the second item. The calculation formula of the cosine similarity can be expressed by the following formula (1) and formula (2):

wherein, W in the formula (1)_ijMay represent a cosine similarity between two items, i.e. a first item and a second item. Wherein the content of the first and second substances,

may indicate the number of users who like item i (e.g., the first item),

may represent a number of users who like item j (e.g., a second item), wherein the heat threshold value

Can be between 0 and 1, wherein,

indicating the need to hit to the maximumThe hotter the door is pressed, i.e., the hotter the item j is, the less similar it is to the item i. It can be understood that the present application adopts a hot punishment rule, that is, the cold user behavior needs to be weighted, and the weight of the active user behavior is reduced, and the specific formula can be referred to the above formula (1). Wherein r in the formula (2)_ikr_jkThe number of users who are liked together can be indicated, for example, the server can count 100 users who like the item i and the item j according to the user information associated with each electronic reading. It is understood that the greater the number of users who are liked in common, the greater the association between the two items, and the greater the similarity. Where n (k) in equation (2) may represent the number of items liked by user k (i.e., the target user), for example, the number of items liked by the target user may be the number of books added to the bookshelf. It can be appreciated that the broader the user interest, the less contribution to the similarity calculation of the two items. It is understood that after the similarity between the first item and the second item in the user terminal shown in fig. 11 is obtained, the top M items ranked in the top may be regarded as the items satisfying the recommendation condition and added to the first recommendation database for storage.

Optionally, if the target dimension information includes second dimension information, a specific process of acquiring the electronic recommended reading material having an association relationship with the electronic collection reading material based on the target dimension information may be described as follows: the server can acquire the classification category to which the electronic collection reading material belongs based on the second dimension information, and pull X data information associated with the classification category to which the electronic collection reading material belongs from an electronic reading material database; x is a positive integer; further, the server may filter the recommended Z data information from the X data information to obtain the remaining Y data information after filtering; y ═ X-Z) and Z is an integer greater than or equal to zero; further, the server can obtain click confidence degrees of the Y data information, and determine the electronic recommended reading materials having an association relation with the electronic collection reading materials from the Y data information according to the click confidence degrees of the Y data information.

For easy understanding, please refer to fig. 12, which is a schematic flow chart illustrating how to make the same kind of good-looking recommendations according to an embodiment of the present application. For a specific implementation manner of the target user performing step S21 and step S22 shown in fig. 12, reference may be made to the description of the electronic recommended reading displayed in the user terminal in the embodiment corresponding to fig. 7 or fig. 8, and details will not be further described here.

In addition, the second recommendation database shown in fig. 12 may be used to store the second click model (which is obtained by training the initial click model in step S27) and the user interest characteristics obtained in the training process. Steps S30-S26 shown in FIG. 12 describe how the initial click model is trained offline by constructed positive and negative examples. Specifically, the server can obtain a positive sample and a negative sample for training an initial click model according to user behavior information associated with each electronic reading in the electronic reading database; the positive sample can comprise a click vector pair formed when a click relation exists between a user and the electronic reading; the negative sample can contain an exposure vector pair formed when the user has an exposure relationship between the electronic readings; further, the server may train the initial click model based on the positive sample and the negative sample, determine the trained initial click model as a second click model, and add the second click model to the second recommendation database.

Herein, it is understood that the positive and negative examples may be collectively referred to as examples, and thus the present application may distinguish the positive and negative examples by the label (0 or 1) of the examples. For example, the label of the sample is 0, that is, the negative sample, and the negative sample may include an exposure vector pair formed by the user id and the item id, that is, the negative sample may be a sample in which the user id has exposure to the item id but does not click, that is, there is no click relationship between the user and the electronic readings displayed in the user terminal for exposure. Conversely, the label of the sample is 1, that is, the positive sample, and the positive sample may include a click vector pair formed by the user id and the item id, that is, the positive sample may be a sample in which the user id has exposure to the item id and has a click, in other words, the user has a click relationship with respect to the electronic readings displayed in the user terminal for exposure.

After the click operation is performed on the interface 100b in the embodiment corresponding to fig. 7 in the electronic reading collection column, the classification cards to which all the classification categories belong are displayed in the interface 200b shown in fig. 7, so that the electronic recommended reading in different classification cards can be viewed by sliding up and down in the interface 200b shown in fig. 7. Before the initial click model is trained, the sample characteristics of the samples need to be extracted and constructed, so that the step S27 can be executed based on the sample characteristics, and when the model parameters of the initial click model are converged, the trained second click model is stored in the second recommendation database, so that the second click model can be accurately sequenced according to the interests of the user in the data recommendation process, the data recommendation accuracy is improved, and the book reading conversion rate of the user can be improved. The sample features may specifically include content-class features (e.g., basic features of the book, statistical-class features such as reading features, exposure features, and scoring features of the book) in step S28, user-class features (e.g., basic representation of the user, behavior statistics of the user), content distribution-class features (e.g., distribution of the book under the interest of age), and word-vector features (e.g., features such as low-dimensional dense vectors of user behaviors and book contents, where the word-vector features may include features of the above-mentioned item vectors).

Optionally, if the target dimension information includes third dimension information, the obtaining, by the server, of the electronic recommended reading that has an association relationship with the electronic collection reading based on the target dimension information may be specifically described as: the server can pull target book information associated with the electronic collection from a third recommendation database based on the third dimension information; further, the server may use the pulled target book form information as an electronic recommended reading having an association relationship with the electronic collection reading.

For easy understanding, please refer to fig. 13, which is a schematic flow chart illustrating a recommendation of a book order according to an embodiment of the present application. For a specific implementation manner of steps S31-S32 shown in fig. 13, reference may be made to the description of the target book entry information in the embodiment corresponding to fig. 9a, and details will not be further described here. It is to be understood that the third recommendation database shown in fig. 13 may be used to store the book information with the similar theme style as the books in the bookshelf, the book information is determined by the server through the document theme generation model (i.e., LDA model) shown in fig. 13, and the book information with the similar theme style may be the second book information. Optionally, step S33 shown in fig. 13 indicates that the third recommendation database may also be used to store operation manuals pre-constructed by an operator according to operation rules, where the operation manuals may include books in a bookshelf, and therefore, in the embodiment of the present application, the information of the manuals where the books in the bookshelf are located may be referred to as first booklet information. It is understood that the search priority of the first order information is higher than the search priority of the second order information based on the step S33.

Optionally, if the target dimension information includes fourth dimension information, a specific process of the server obtaining the electronic recommended reading material having an association relationship with the electronic collection reading material based on the target dimension information may be described as follows: acquiring the electronic collection reading materials carrying the reading completion identification from the electronic reading material collection column based on the fourth dimension information; acquiring a completion timestamp of the electronic collection reading materials carrying the read-out identification, sequencing the electronic collection reading materials carrying the read-out identification based on the completion timestamp, and determining the sequenced electronic collection reading materials carrying the read-out identification as the finished electronic reading materials; and taking the completed electronic reading material and the comment information of the completed electronic reading material as the electronic recommended reading material which has an association relation with the electronic collection reading material.

For easy understanding, please refer to fig. 14, which is a schematic flowchart illustrating a recommendation completing process according to an embodiment of the present application. For a specific implementation manner of steps S41-S42 shown in fig. 14, reference may be made to the specific description of the electronic collection reading with reading identifier in the embodiment corresponding to fig. 9b, and details will not be further described here. It can be understood that, in the embodiment of the present application, the electronic collections in the bookshelf that carry the read-out identifiers may also be stored in the fourth recommendation database in advance, so that when the data pull request is received, the electronic collections in the fourth recommendation database that carry the read-out identifiers may be quickly sorted according to the order of the completion timestamp (for example, the electronic collections that carry the read-out identifiers may be sorted based on the near-far sorting rule shown in fig. 14), and thus the sorted electronic collections that carry the read-out identifiers may be collectively referred to as completed electronic collections. It is understood that the fourth recommendation database can also be used for storing the comment information of the finished electronic readings in advance. At this time, taking the number of books read by the target user in the bookshelf as book a as an example, the server may pull the comment information of the book a, and quickly find out the comment information posted by the target user from the comment information as the comment information of the completed electronic reading, and on the contrary, may use the comment information with the highest praise number from the comment information as the comment information of the completed electronic reading.

Therefore, when a target user performs a click operation on an electronic reading collection column (for example, a collection interface in the electronic reading collection column or a data push interface associated with the electronic reading collection column), the user terminal may add dimension information (i.e., target dimension information) corresponding to the click operation to a data pull request, and send the data pull request carrying the target dimension information to the server associated with the reading application (i.e., the aforementioned QQ reading application), so that the server may obtain the target dimension information carried in the book request instruction, and may obtain a data push list associated with the target dimension information, and may display an electronic recommended reading having an association relationship with the electronic reading in the data push list associated with the target dimension information. Therefore, through multi-dimensional data recommendation in the application scene of the electronic bookshelf, the interests and hobbies of the user can be fully mined, and then the electronic books fitting the interests and hobbies of the user can be efficiently and accurately matched, so that the book searching efficiency and the book searching accuracy of the user in the bookshelf are improved.

Further, please refer to fig. 15, which is a schematic structural diagram of an electronic reading data processing apparatus according to an embodiment of the present application. The electronic reading material data processing device 1 can be applied to the user terminal, which can be the user terminal 300a in the embodiment corresponding to fig. 1. Further, the electronic reading data processing apparatus 1 may include: the system comprises a dimension information determining module 100, a dimension information sending module 200 and a recommendation list acquiring module 300;

the dimension information determining module 100 is configured to respond to a first operation triggered for an electronic reading collection column, and determine, in multiple dimension information corresponding to the electronic reading collection column, the dimension information corresponding to the first operation as target dimension information; the electronic reading matter collecting column comprises electronic collecting reading matters;

a dimension information sending module 200, configured to send the target dimension information to a server;

a recommendation list obtaining module 300, configured to obtain a data push list associated with the target dimension information and returned by the server; the data push list comprises electronic recommended reading materials which are in association relation with the electronic collection reading materials.

For specific implementation manners of the dimension information determining module 100, the dimension information sending module 200, and the recommendation list obtaining module 300, reference may be made to the description of the user terminal in the embodiment corresponding to fig. 3, which will not be described again.

Further, please refer to fig. 16, which is a schematic structural diagram of a computer device according to an embodiment of the present application. As shown in fig. 16, the computer device 1000 may be applied to a user terminal, which may be the user terminal 3000a in the embodiment corresponding to fig. 1. The computer device 1000 may include: the processor 1001, the network interface 1004, and the memory 1005, the computer apparatus 1000 may further include: a user interface 1003, and at least one communication bus 1002. Wherein a communication bus 1002 is used to enable connective communication between these components. The user interface 1003 may include a Display screen (Display) and a Keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface and a standard wireless interface. The network interface 1004 may optionally include a standard wired interface, a wireless interface (e.g., WI-FI interface). The memory 1004 may be a high-speed RAM memory or a non-volatile memory (e.g., at least one disk memory). The memory 1005 may optionally be at least one memory device located remotely from the processor 1001. As shown in fig. 16, a memory 1005, which is a kind of computer storage medium, may include therein an operating system, a network communication module, a user interface module, and a device control application program.

The network interface 1004 in the computer device 1000 may also be connected to the server in the embodiment corresponding to fig. 1 through a network, and the optional user interface 1003 may further include a Display screen (Display) and a Keyboard (Keyboard). In the computer device 1000 shown in fig. 16, the network interface 1004 may provide a network communication function; the user interface 1003 is an interface for providing a user with input; and the processor 1001 may be used to invoke a device control application stored in the memory 1005 to implement:

sending the target dimension information to a server;

It should be understood that the computer device 1000 described in this embodiment of the present application can perform the description of the user terminal in the embodiment corresponding to fig. 3, and can also perform the description of the electronic reading data processing apparatus 1 in the embodiment corresponding to fig. 15, which is not described herein again. In addition, the beneficial effects of the same method are not described in detail.

Further, here, it is to be noted that: an embodiment of the present application further provides a computer storage medium, and the computer storage medium stores the aforementioned computer program executed by the electronic reading data processing apparatus 1, and the computer program includes program instructions, and when the processor executes the program instructions, the description of the electronic reading data processing method in the embodiment corresponding to fig. 3 can be performed, so that details are not repeated here. In addition, the beneficial effects of the same method are not described in detail. For technical details not disclosed in the embodiments of the computer storage medium referred to in the present application, reference is made to the description of the embodiments of the method of the present application.

Further, please refer to fig. 17, which is a schematic structural diagram of another electronic reading data processing apparatus according to an embodiment of the present application. The electronic reading data processing device 2 can be applied to the server, which can be the service server 2000 in the embodiment corresponding to fig. 1. Further, the electronic reading data processing device 2 may include: the system comprises a dimension information acquisition module 10, a reading material recommendation module 20 and a recommendation list generation module 30;

the dimension information acquisition module 10 is used for acquiring target dimension information sent by a user terminal; the target dimension information is obtained by a user terminal responding to a first operation aiming at a plurality of dimension information corresponding to the electronic reading collection column; the electronic reading matter collecting column comprises electronic collecting reading matters;

the reading recommending module 20 is configured to obtain an electronic recommended reading having an association relationship with the electronic collection reading based on the target dimension information;

wherein the target dimension information comprises first dimension information;

the reading recommendation module 20 comprises: a first pulling unit 201, a first filtering unit 202; further, the reading recommendation module further comprises: an item vector constructing unit 203, a first item determining unit 204, a second item determining unit 205, a similarity determining unit 206;

a first pulling unit 201, configured to pull N pieces of data information matching the electronic collection from a first recommendation database based on the first dimension information; n is a positive integer;

wherein the first pulling unit 201 includes: a model acquisition subunit 2011, a candidate determination subunit 2012, a probability value determination subunit 2013, and an ordering subunit 2014;

a model obtaining subunit 2011, configured to obtain, based on the first dimension information, a first click model associated with the first item;

a candidate determining subunit 2012, configured to determine the data information in the first recommendation database as M candidate data information; m is the total amount of data information in the recommendation database;

a probability value determination subunit 2013, configured to determine predicted probability values of the M candidate data information based on the first click model;

the sorting subunit 2014 is configured to sort the prediction probability values of the M pieces of candidate data information, and select N pieces of data information matched with the electronic collection from the sorted M pieces of candidate data information based on the prediction probability values of the sorted M pieces of candidate data information; n is less than or equal to M.

For a specific implementation manner of the model obtaining subunit 2011, the candidate determining subunit 2012, the probability value determining subunit 2013, and the ordering subunit 2014, reference may be made to the description of the N pieces of data information in the embodiment corresponding to fig. 10, which will not be described again here.

The first filtering unit 202 is configured to filter S pieces of recommended data information from the N pieces of data information, and use K pieces of data information remaining after filtering as an electronic recommended reading material having an association relationship with the electronic collection reading material; k ═ N-S and S is an integer greater than or equal to zero.

Optionally, the item vector constructing unit 203 is configured to obtain an item vector of each electronic reading according to a keyword carried by each electronic reading in the electronic reading database and user information associated with each electronic reading;

wherein, the item vector construction unit 203 comprises: an information acquisition subunit 2031, an information analysis subunit 2032, a word extraction subunit 2033, and a vector construction subunit 2034;

an information acquiring subunit 2031, configured to acquire all electronic readings in the electronic reading database and user behavior information associated with all electronic readings;

the information analyzing subunit 2032 is configured to analyze the user behavior information to obtain user information associated with each electronic reading, and obtain a user sequence of each electronic reading according to the user information associated with each electronic reading;

a word extracting subunit 2033, configured to extract a keyword carried by each electronic reading, and obtain a keyword sequence of each electronic reading according to the keyword of each electronic reading;

a vector construction subunit 2034, configured to construct an item vector of each electronic reading material based on the user sequence of each electronic reading material and the keyword sequence of each electronic reading material.

The specific implementation manners of the information obtaining subunit 2031, the information analyzing subunit 2032, the word extracting subunit 2033, and the vector constructing subunit 2034 may refer to the description of the article vector in the embodiment corresponding to fig. 10, which will not be described again.

A first item determining unit 204, configured to determine the electronic collection reading in the electronic reading database as a first item, and obtain an item vector of the first item;

a second item determining unit 205, configured to determine, as a second item, an electronic reading in the electronic reading database except for the electronic collection reading, so as to obtain an item vector of the second item;

the similarity determining unit 206 is configured to determine cosine similarity between the item vector of the first item and the item vector of the second item, use the second item of which the cosine similarity meets a recommendation condition as data information matched with the electronic collection reading, and add the data information matched with the electronic collection reading to the first recommendation database.

For specific implementation manners of the first pulling unit 201, the first filtering unit 202, the item vector constructing unit 203, the first item determining unit 204, the second item determining unit 205, and the similarity determining unit 206, reference may be made to the description of the first dimension information in the embodiment corresponding to fig. 10, which will not be described again.

Optionally, the target dimension information includes second dimension information;

the reading recommendation module 20 can include: a second pulling unit 207, a second filtering unit 208, a sorting unit 209; further, the reading recommendation module 20 further comprises: a sample determination unit 210, a model training unit 211;

a second pulling unit 207, configured to obtain a classification category to which the electronic collection reading belongs based on the second dimension information, and pull X pieces of data information associated with the classification category to which the electronic collection reading belongs from an electronic reading database; x is a positive integer;

the second filtering unit 208 is configured to filter the recommended Z data information from the X data information to obtain the remaining Y data information after filtering; y ═ X-Z) and Z is an integer greater than or equal to zero;

and the sorting unit 209 is configured to acquire the click confidence degrees of the Y pieces of data information, and determine an electronic recommended reading having an association relationship with the electronic collection reading from the Y pieces of data information according to the click confidence degrees of the Y pieces of data information.

Wherein the sorting unit 209 includes: a confidence level determination subunit 2091, a data selection subunit 2092;

the confidence degree determining subunit 2091 is configured to obtain a second click model from a second recommendation database, determine the click confidence degree of the Y data information based on the second click model and the user interest feature, and sort the Y data information according to the click confidence degree of the Y data information;

a data selecting subunit 2092, configured to select K pieces of data information from the sorted Y pieces of data information as an electronic recommended reading having an association relationship with the electronic collection reading; and K is an integer less than or equal to Y.

The specific implementation manners of the confidence determining subunit 2091 and the data selecting subunit 2092 may refer to the description of the second click model in the embodiment corresponding to fig. 10, and details will not be further described here.

Optionally, the sample determining unit 210 is configured to obtain a positive sample and a negative sample for training an initial click model according to user behavior information associated with each electronic reading in the electronic reading database; the positive sample comprises a click vector pair formed when a click relation exists between a user and the electronic reading; the negative sample comprises an exposure vector pair formed when the user has an exposure relation with the electronic reading material;

and the model training unit 211 is configured to train the initial click model based on the positive sample and the negative sample, determine the trained initial click model as a second click model, and add the second click model to the second recommendation database.

For specific implementation manners of the second pulling unit 207, the second filtering unit 208, the sorting unit 209, the sample determining unit 210, and the model training unit 211, reference may be made to the description of the second dimension information in the embodiment corresponding to fig. 10, which will not be described again here.

Optionally, the target dimension information includes third dimension information;

the reading recommendation module 20 can include: a third pull unit 212, a reading material determination unit 213; further, the reading recommendation module 20 further comprises: a theme analysis unit 214, a cluster dividing unit 215, a book form generation unit 216, and a book form adding unit 217;

a third pulling unit 212, configured to pull target book form information associated with the electronic collection from a third recommendation database based on the third dimension information;

wherein the third pulling unit 212 includes: a first search subunit 2121 and a book order determination subunit 2122;

a first search sub-list 2121, configured to preferentially search, in a recommendation database, first book list information including the electronic collection reading materials based on the third dimension information;

a book form determining subunit 2122, configured to, if first book form information including the electronic collection reading material is searched, use the first book form information as target book form information associated with the electronic collection reading material;

the book form determining subunit 2122 is further configured to search, if the first book form information including the electronic collection reading material is not searched, second book form information associated with the subject information of the electronic collection reading material from the recommendation database, and use the searched second book form information as target book form information associated with the electronic collection reading material; the search priority of the first order information is higher than the search priority of the second order information.

For specific implementation manners of the first searching subunit 2121 and the book list determining subunit 2122, reference may be made to the description of the target book list information in the embodiment corresponding to fig. 10, and details will not be described here again.

And the reading material determining unit 213 is configured to use the pulled target book form information as an electronic recommended reading material having an association relationship with the electronic collection reading material.

The theme analysis unit 214 is configured to perform theme analysis on the theme information of all the electronic readings in the electronic reading database through a theme analysis model to obtain the theme information of each electronic reading;

a cluster dividing unit 215 for dividing the cluster to which the subject information of each electronic reading belongs; one cluster corresponds to one topic information;

a book list generating unit 216, configured to search a cluster associated with the topic information of the electronic collection reading in the cluster, generate second book list information according to the searched electronic reading in the cluster, and add the second book list information to the third recommendation database;

a book form adding unit 217, configured to add the first book form information created based on the operation rule to the third recommendation database.

For specific implementation manners of the third pulling unit 212, the reading material determining unit 213, the topic analyzing unit 214, the cluster dividing unit 215, the book list generating unit 216, and the book list adding unit 217, reference may be made to the description of the third dimension information in the embodiment corresponding to fig. 10, which will not be described again.

Optionally, the target dimension information includes fourth dimension information;

the reading recommendation module 20 can include: a read completion acquisition unit 218, a timestamp sorting unit 219, and a comment determination unit 220;

a read-out acquisition unit 218, configured to acquire an electronic collection reading carrying a read-out identifier from the electronic reading collection column based on the fourth dimension information;

the time stamp sorting unit 219 is configured to acquire a completion time stamp of the electronic collection reading with the read-out completion identifier, sort the electronic collection reading with the read-out completion identifier based on the completion time stamp, and determine the sorted electronic collection reading with the read-out completion identifier as a completed electronic reading;

and a comment determining unit 220, configured to use the completed electronic reading and comment information of the completed electronic reading as an electronic recommended reading having an association relationship with the electronic collection.

For specific implementation manners of the read-out completion obtaining unit 218, the timestamp sorting unit 219, and the comment determining unit 220, reference may be made to the description of the fourth dimension information in the embodiment corresponding to fig. 10, which will not be described again.

And the recommendation list generating module 30 is configured to generate a data push list associated with the target dimension information according to the electronic recommended reading material, and return the data push list to the user terminal.

For specific implementation manners of the dimension information obtaining module 10, the reading material recommending module 20, and the recommendation list generating module 30, reference may be made to the description of the server in the embodiment corresponding to fig. 3, and details will not be further described here.

Further, please refer to fig. 18, which is a schematic structural diagram of another computer device according to an embodiment of the present application. As shown in fig. 18, the computer device 2000 may be applied to a server, which may be the service server 2000 in the embodiment corresponding to fig. 1. The computer device 2000 may include: a processor 2001, a network interface 2004, and a memory 2005, the computer device may further include: a user interface 2003, and at least one communication bus 2002. The communication bus 2002 is used to implement connection communication between these components. The user interface 2003 may include a Display (Display) and a Keyboard (Keyboard), and the optional user interface 2003 may further include a standard wired interface and a standard wireless interface. The network interface 2004 may optionally include a standard wired interface, a wireless interface (e.g., WI-FI interface). The memory 2004 may be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. The memory 2005 may optionally also be at least one memory device located remotely from the aforementioned processor 2001. As shown in fig. 18, the memory 2005, which is one type of computer storage medium, may include therein an operating system, a network communication module, a user interface module, and a device control application program.

The network interface 2004 in the computer device 2000 may also be connected to the user terminal in the embodiment corresponding to fig. 1, and the optional user interface 2003 may further include a Display screen (Display) and a Keyboard (Keyboard). In the computer device 2000 shown in fig. 18, the network interface 2004 may provide a network communication function; and the user interface 2003 is primarily used to provide an interface for user input; and processor 2001 may be used to invoke the device control application stored in memory 2005 to implement:

It should be understood that the computer device 2000 described in this embodiment of the application can perform the description of the electronic reading data processing method in the embodiment corresponding to fig. 3 or fig. 10, and can also perform the description of the electronic reading data processing apparatus 2 in the embodiment corresponding to fig. 17, which is not repeated herein. In addition, the beneficial effects of the same method are not described in detail.

Further, here, it is to be noted that: an embodiment of the present application further provides a computer storage medium, and the computer storage medium stores the aforementioned computer program executed by the electronic reading data processing apparatus 2, and the computer program includes program instructions, and when the processor executes the program instructions, the description of the electronic reading data processing method in the embodiment corresponding to fig. 3 or fig. 10 can be executed, so that details are not repeated here. In addition, the beneficial effects of the same method are not described in detail. For technical details not disclosed in the embodiments of the computer storage medium referred to in the present application, reference is made to the description of the embodiments of the method of the present application.

Further, please refer to fig. 19, which shows an electronic reading data processing system according to an embodiment of the present application. The electronic reading material data processing system 3 may include a user terminal 1 and a server 2, where the user terminal 1 may be the image data processing apparatus 1 in the embodiment corresponding to fig. 15; the server may be the image data processing apparatus 2 in the embodiment corresponding to fig. 17. It is understood that the beneficial effects of the same method are not described in detail.

It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.

The above disclosure is only for the purpose of illustrating the preferred embodiments of the present application and is not to be construed as limiting the scope of the present application, so that the present application is not limited thereto, and all equivalent variations and modifications can be made to the present application.

Claims

1. An electronic reading data processing method, which is applied to a user terminal, is characterized by comprising the following steps:

sending the target dimension information to a server;

2. An electronic reading data processing method, which is applied to a server, is characterized by comprising the following steps:

3. The method of claim 2, wherein the target dimension information comprises first dimension information;

4. The method of claim 3, further comprising:

5. The method according to claim 4, wherein the obtaining of the item vector of each electronic reading material according to the keywords carried by each electronic reading material in the electronic reading material database and the user information associated with each electronic reading material comprises:

6. The method of claim 4, wherein the pulling N data messages matching the electronic collection from a first recommendation database based on the first dimension information comprises:

7. The method of claim 2, wherein the target dimension information comprises second dimension information;

8. The method of claim 7, wherein the obtaining of the click confidence of the Y data information and the determining of the electronic recommended reading material having an association relationship with the electronic collection reading material from the Y data information according to the click confidence of the Y data information comprises:

9. The method of claim 7, further comprising:

10. The method of claim 2, wherein the target dimension information comprises third dimension information;

11. The method of claim 10, wherein pulling target book information associated with the electronic collection from a third recommendation database based on the third dimension information comprises:

12. The method of claim 11, further comprising:

13. The method of claim 2, wherein the target dimension information comprises fourth dimension information;

14. An electronic reading data processing device, which is applied to a user terminal, is characterized by comprising:

15. An electronic reading data processing device, which is applied to a server, is characterized by comprising: