CN111523034A

CN111523034A - Application processing method, device, equipment and medium

Info

Publication number: CN111523034A
Application number: CN202010331435.XA
Authority: CN
Inventors: 梁华盛; 何文
Original assignee: Tencent Technology Shenzhen Co Ltd
Current assignee: Tencent Technology Shenzhen Co Ltd
Priority date: 2020-04-24
Filing date: 2020-04-24
Publication date: 2020-08-11
Anticipated expiration: 2040-04-24
Also published as: CN111523034B

Abstract

The embodiment of the application discloses a processing method of an application program. The method comprises the following steps: and acquiring a second application program from the reference program set according to the main body attribute of the first application program, wherein the main body attribute of the second application program is associated with the main body attribute of the first application program, predicting the quality characteristic of the first application program by adopting the quality characteristic of the second application program, and outputting the quality identification result of the first application program according to the quality characteristic of the first application program. Based on the idea of clustering, a second application program which is the same as the first application program is mined through the relevance between the main attributes, the quality characteristics of the first application program, which are obtained by adopting the quality characteristic prediction of the second application program, can reflect the overall quality of the first application program more accurately, and the more accurate quality identification result of the first application program is obtained based on the quality characteristics of the first application program.

Description

Application processing method, device, equipment and medium

Technical Field

The present application relates to the field of computer technologies, and in particular, to the field of application-related technologies, and in particular, to an application processing method, an application processing apparatus, an application processing device, and a computer-readable storage medium.

Background

With the development of computer technology, an installation-free application (i.e., an application that can be used without downloading and installing) is developed rapidly and rapidly with the advantages of convenience and small memory occupation. The application programs have the characteristics of multiple categories, complex functions, various carriers, multiple content, multiple structures and the like, so that the quality of the application programs is difficult to be well identified by using a single technology. At present, two quality identification technologies for the application program are mainly used, one is manual review, and special operation and maintenance personnel review the content, functions and the like of the application program to remove some low-quality application programs; this approach is inefficient, costly and not practical. The other method is to adopt the abnormal matching of the model, specifically to adopt some low-quality data to carry out modeling, match the content, the function and the like related to the application program with the model, and if the matching is successful, the content and the function of the application program are regarded as the low-quality data, so as to identify the quality abnormality of the application program; the method has the problems of low precision, difficulty in integrating complex information such as functions, pages, interaction and the like in the application program, low efficiency and high cost.

Disclosure of Invention

The embodiment of the application discloses a processing method, a processing device, processing equipment and a processing medium of an application program, which can accurately identify the quality of the application program through a simple and efficient quality identification technology.

In one aspect, an embodiment of the present application provides a method for processing an application program, where the method includes:

acquiring the main body attribute and a reference program set of a first application program, wherein the reference program set comprises the main body attribute and the quality characteristic of a plurality of reference application programs;

obtaining a second application program from the reference program set, wherein the second application program is at least one reference application program in the reference program set, and the body attribute of the second application program is associated with the body attribute of the first application program;

predicting the quality characteristics of the first application program by adopting the quality characteristics of the second application program;

and outputting a quality identification result of the first application program according to the quality characteristics of the first application program.

In the embodiment of the application, a second application program is acquired from a reference program set according to the main body attribute of the first application program, the main body attribute of the second application program is associated with the main body attribute of the first application program, the quality characteristic of the first application program is predicted by adopting the quality characteristic of the second application program, and the quality identification result of the first application program is output according to the quality characteristic of the first application program. Based on the idea of clustering, a second application program which is the same as the first application program is mined through the relevance between the main attributes, the quality characteristics of the first application program, which are obtained by adopting the quality characteristic prediction of the second application program, can reflect the overall quality of the first application program more accurately, and the more accurate quality identification result of the first application program is obtained based on the quality characteristics of the first application program.

displaying a search page of a first client, wherein the first client comprises a first application program, and the first application program refers to any sub-application program parasitic in the first client;

acquiring a quality identification result of the first application program, wherein the quality identification result of the first application program is obtained by adopting the processing method of the application program;

if the quality identification result of the first application program meets the recommendation condition, acquiring information of the first application program;

information of the first application is displayed in a search page of the first client.

In the embodiment of the application, a search page of the first client is displayed, a quality identification result of the first application program is obtained, the quality identification result of the first application program is obtained by adopting the processing method of the application program, if the quality identification result of the first application program meets the recommendation condition, information of the first application program is obtained, and the information of the first application program is displayed in the search page of the first client. Therefore, the application programs meeting the recommendation conditions are recommended, the probability that the high-quality application programs are selected and used by the user is improved, and the user experience is further improved.

In one aspect, the present application provides an application processing apparatus, including:

the device comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring the main body attribute and a reference program set of a first application program, and the reference program set comprises the main body attribute and the quality characteristic of a plurality of reference application programs;

and the processing unit is used for acquiring a second application program from the reference program set, wherein the second application program is at least one reference application program in the reference program set, the main body attribute of the second application program is associated with the main body attribute of the first application program, the quality characteristic of the first application program is predicted by adopting the quality characteristic of the second application program, and the quality identification result of the first application program is output according to the quality characteristic of the first application program.

In an embodiment, the processing unit is further configured to obtain a second application from the reference program set, and specifically to:

constructing a main body relational graph, wherein the main body relational graph comprises attribute nodes and program nodes, the attribute nodes comprise main body attribute nodes of a first application program and main body attribute nodes of a reference application program, the program nodes comprise the first application program nodes and the reference application program nodes, and one attribute node is connected with a plurality of program nodes;

splitting the main body relationship graph according to the attribute nodes to obtain a plurality of sub relationship graphs;

acquiring a target sub-relational graph from the plurality of sub-relational graphs, and determining a second application program from the target sub-relational graph; the attribute nodes of the target sub-relational graph are subject attribute nodes of the first application program, and the program nodes of the target sub-relational graph comprise the first application program node and the second application program node.

In an embodiment, the processing unit is further configured to predict the quality characteristic of the first application using the quality characteristic of the second application, and specifically to:

constructing a characteristic diagram of the first application program according to the quality characteristics of the second application program and the target sub-relational diagram;

and calling the graph network model to perform prediction processing on the feature graph to obtain the quality feature of the first application program.

In an embodiment, the processing unit is further configured to construct a feature map of the first application according to the quality feature of the second application and the target sub-relationship map, and specifically configured to:

connecting program nodes in the target sub-relational graph with each other, and deleting attribute nodes in the target sub-relational graph to obtain a characteristic graph of the first application program;

the program nodes in the characteristic diagram comprise a first application program node and a second application program node, the second application program node carries the quality characteristics of the second application program, and an edge formed by connecting any two program nodes carries the main body attribute of the first application program.

In one embodiment, the graph network model includes a graph convolution network;

the processing unit is further configured to call the graph network model to perform prediction processing on the feature graph to obtain a quality feature of the first application program, and specifically configured to:

and calling the graph convolutional network to extract the quality characteristics of the first application program from the characteristic graph.

In one embodiment, the number of the main attributes of the first application program is N, the number of the feature maps of the first application program is N, the graph network model includes a graph convolution network, the graph convolution network includes N layers of graph convolution layers, and N is an integer greater than 1;

calling N layers of graph convolution layers to respectively extract N quality characteristic components of the first application program from the N characteristic graphs; wherein, a layer of graph convolution layer is used for carrying out feature extraction processing on a feature graph to obtain a quality feature component of the first application program;

performing aggregation processing on the N quality characteristic components to obtain the quality characteristics of the first application program;

the aggregation processing mode comprises a weighted aggregation mode or a concatenation mode.

calling a graph convolution network to carry out convolution processing on the feature graph to obtain an intermediate quality feature;

and stacking the intermediate quality characteristics to obtain the quality characteristics of the first application program.

In one embodiment, the graph network model includes a classifier;

the processing unit is further configured to output a quality identification result of the first application program according to the quality characteristic of the first application program, and specifically configured to:

and calling a classifier to classify the quality characteristics of the first application program to obtain a quality identification result of the first application program.

In one embodiment, the processing unit is further configured to:

adding the subject attributes and quality features of the first application to a reference program set;

wherein the subject attributes include at least one of: an identification of the developer, an identification of the operator, and a service address;

the first application and the reference application are both installation-free applications; the first application program refers to any sub-application program parasitic in the first client side; the reference application program refers to a sub-application program which is parasitized in the first client side and is except the first application program; alternatively, the reference application refers to a sub-application hosted within the second client.

the system comprises a display unit, a search unit and a search unit, wherein the display unit is used for displaying a search page of a first client, the first client comprises a first application program, and the first application program refers to any sub-application program parasitic in the first client;

the processing unit is used for acquiring a quality identification result of the first application program, wherein the quality identification result of the first application program is obtained by adopting the processing method of the application program; if the quality identification result of the first application program meets the recommendation condition, acquiring information of the first application program;

and the display unit is also used for displaying the information of the first application program in the search page of the first client.

In one embodiment, the quality identification result includes a quality score; the quality identification result of the first application program meets the recommendation condition, namely the quality score of the first application program is larger than the quality threshold; the information of the first application includes at least one of: identification, subject attributes, categories, function profiles of the first application;

the processing unit is further configured to display information of the first application program in a search page of the first client, and specifically configured to:

adding information of a first application program into a recommendation list of a first client, wherein the recommendation list comprises a plurality of pieces of information to be recommended, and the information is sorted according to the sequence of the quality scores of the corresponding application programs from high to low;

displaying information in a recommendation list of a first client in a search page of the first client;

the processing unit is further configured to:

if the quality score of the first application program is smaller than a penalty threshold, performing penalty processing on the first application program, wherein the penalty threshold is smaller than the quality threshold, and the penalty processing comprises at least one of the following steps: and shielding, deleting and sending the modification prompt information to the operator of the first application program.

In one aspect, the present application provides an application processing device, including:

a memory storing computer readable instructions;

a processor coupled to the memory for executing the computer readable instructions to implement the processing method of the application program described above.

In one aspect, the present application provides a computer-readable storage medium storing one or more instructions adapted to be loaded by a processor and to execute the processing method of the application program.

In the embodiment of the application, a second application program is acquired from a reference program set according to the main body attribute of the first application program, the main body attribute of the second application program is associated with the main body attribute of the first application program, the quality characteristic of the first application program is predicted by adopting the quality characteristic of the second application program, and the quality identification result of the first application program is output according to the quality characteristic of the first application program. Based on the idea of clustering, a second application program which is the same as the first application program is mined through the relevance between the main attributes, the quality characteristics of the first application program, which are obtained through the quality characteristic prediction of the second application program, can reflect the overall quality of the first application program more accurately, and the quality identification technology is simple, efficient, low in cost consumption and higher in practicability through obtaining the more accurate quality identification result of the first application program based on the quality characteristics of the first application program; when the first client side searches the application programs, if the quality identification result of the first application program shows that the quality of the first application program is better, the first application program can be recommended in the searching process, the probability that the high-quality application program is selected and used by a user is improved, and further the user experience is improved.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

FIG. 1 illustrates an architecture diagram of a processing system for an application provided by an exemplary embodiment of the present application;

FIG. 2 is a flow chart illustrating a method for processing an application according to an exemplary embodiment of the present application;

FIG. 3 is a flowchart illustrating a processing method for an application according to an exemplary embodiment of the present application;

FIG. 4a illustrates a body relationship diagram provided by an exemplary embodiment of the present application;

FIG. 4b illustrates a sub-relationship diagram provided by an exemplary embodiment of the present application;

FIG. 4c illustrates a feature diagram provided by an exemplary embodiment of the present application;

FIG. 4d illustrates a flowchart for extracting quality features of an application in a feature map provided by an exemplary embodiment of the present application;

FIG. 4e is a flow chart illustrating a further method for extracting quality features of an application in a feature map according to an exemplary embodiment of the present application;

FIG. 4f illustrates a flow chart for extracting quality features of an application in a feature map according to an exemplary embodiment of the present application;

FIG. 4g illustrates yet another body relationship diagram provided by an exemplary embodiment of the present application;

FIG. 4h illustrates a flowchart of another method for extracting quality features of an application in a feature map provided by an exemplary embodiment of the present application;

FIG. 4i illustrates a flowchart of another method for extracting quality features of an application in a feature map provided by an exemplary embodiment of the present application;

FIG. 5 is a flow chart illustrating a method for processing a further application provided by an exemplary embodiment of the present application;

FIG. 6a illustrates a search page diagram of a first client provided by an exemplary embodiment of the present application;

FIG. 6b illustrates a search results page diagram of a first client provided by an exemplary embodiment of the present application;

FIG. 7 is a flowchart illustrating a method for processing another application program according to an exemplary embodiment of the present application;

fig. 8 is a schematic structural diagram illustrating a processing apparatus of an application according to an exemplary embodiment of the present application;

FIG. 9 is a schematic diagram illustrating an exemplary embodiment of a processing device for processing a further application;

fig. 10 is a schematic structural diagram illustrating a processing device of an application according to an exemplary embodiment of the present application.

Detailed Description

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

The embodiment of the application relates to Artificial Intelligence (AI), Natural Language Processing (NLP) and Machine Learning (ML), and hidden information with potential value in data can be mined by combining the AI, the NLP and the ML, so that equipment can predict and identify an application more accurately. The AI is a theory, method, technique and application system that simulates, extends and expands human intelligence, senses the environment, acquires knowledge and uses the knowledge to obtain the best results using a digital computer or a machine controlled by a digital computer. In other words, artificial intelligence is a comprehensive technique of computer science that attempts to understand the essence of intelligence and produce a new intelligent machine that can react in a manner similar to human intelligence. Artificial intelligence is the research of the design principle and the realization method of various intelligent machines, so that the machines have the functions of perception, reasoning and decision making.

The AI technology is a comprehensive subject, and relates to the field of extensive technology, both hardware level technology and software level technology. The artificial intelligence infrastructure generally includes technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, processing technologies for large applications, operating/interactive systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a voice processing technology, a natural language processing technology, machine learning/deep learning and the like.

NLP is an important direction in the fields of computer science and AI. It studies various theories and methods that enable efficient communication between humans and computers using natural language. NLP is a science integrating linguistics, computer science and mathematics. Therefore, the research in this field will involve natural language, i.e. the language that people use everyday, so it is closely related to the research of linguistics. NLP techniques typically include text processing, semantic understanding, machine translation, robotic question and answer, knowledge mapping, and the like.

ML is a multi-field interdisciplinary, and relates to a plurality of disciplines such as probability theory, statistics, approximation theory, convex analysis, algorithm complexity theory and the like. The special research on how a computer simulates or realizes the learning behavior of human beings so as to acquire new knowledge or skills and reorganize the existing knowledge structure to continuously improve the performance of the computer. ML is the core of artificial intelligence, is the fundamental way to make computers intelligent, and its application is spread over various fields of artificial intelligence. ML and deep learning generally include techniques such as artificial neural networks, belief networks, reinforcement learning, migratory learning, inductive learning, and formal learning.

The embodiment of the application relates to processing of an application program. An application herein may refer to an installation-free application, i.e., an application that can be used without downloading installation, which is also known as an applet, and is typically hosted in a client as a sub-application. The client (which may also be referred to as an application client, APP client) refers to a program that is downloaded and installed in the terminal and runs in the terminal. Various types of clients may be included in the terminal, including but not limited to: an IM (instant messaging) client (e.g., a wechat client, a QQ client, etc.), an SNS (social networking Services) client (e.g., a microblog client, a map client with social networking functions, etc.), a content service client (e.g., a news client), an image processing client, a search client, etc. Unless otherwise noted, the application programs mentioned in the following embodiments of the present application are all described by taking sub-application programs (i.e., applets) hosted by various clients of the terminal as examples.

Due to the wide variety and quantity of the application programs, the quality of the application programs is also uneven, and the application programs can be generally divided into high-quality application programs and low-quality application programs. Wherein the quality of the application can be evaluated from multiple dimensions of traffic, services, content, etc. From the traffic dimension, a good application may refer to an application whose traffic (e.g., accumulated praise or attention) exceeds a threshold, for example: if the accumulated number of praise of a user to an application exceeds 100 ten thousand, the application can be regarded as a good application. Conversely, a bad application is an application that does not reach a threshold value for flow, or an application that is reported to increase flow by an improper means, such as: applications that increase traffic by falsifying information such as names and trademarks of others or by inducing sharing and purchasing false attention are considered to be inferior applications. From the service dimension, the high-quality application may refer to an application that can smoothly provide a service for a user, and the low-quality application may refer to an application that cannot normally provide a service for a user, for example: applications that provide unavailable services due to expired and abnormal certificates or malicious services such as induced downloads and plug-ins are considered as inferior applications. From the content dimension, the priority application program refers to an application program which does not contain advertisement information, has positive energy and does not contain illegal or false information, and the content of the application program is positive and upward; the inferior application program refers to an application program with a large amount of advertisement information or illegal website skip links in the content. Therefore, identifying the quality of the application programs and screening out high-quality application programs to provide high-quality services for users are important means for improving client competitiveness. However, there are major challenges to quality identification of applications because: different application programs have different functions; for example: the news applet provides news searching and browsing functions, and the shopping applet provides an e-commerce function; the content of the application program is rich and various, for example, the news applet comprises various news; the shopping applet contains various commodity information; thirdly, the carriers of the application programs are various; for example, a shopping applet is hosted by an IM client, while a game applet is hosted by a news client; the structures of the application programs are different, and the application programs are developed by using different program frameworks; for the above reasons, there are major technical challenges to the quality identification of applications.

The embodiment of the application provides a scheme for identifying and processing the quality of an application program, the scheme acquires a second application program from a reference program set according to a main attribute of the first application program, the main attribute of the second application program is associated with the main attribute of the first application program, the quality characteristic of the first application program is predicted by adopting the quality characteristic of the second application program, and the quality identification result of the first application program is output according to the quality characteristic of the first application program. Based on the idea of clustering, a second application program which is the same as the first application program is mined through the relevance between the main attributes, the quality characteristics of the first application program, which are obtained through the quality characteristic prediction of the second application program, can reflect the overall quality of the first application program more accurately, and the quality identification technology is simple, efficient, low in cost consumption and higher in practicability through obtaining the more accurate quality identification result of the first application program based on the quality characteristics of the first application program; when the first client side searches the application programs, if the quality identification result of the first application program shows that the quality of the first application program is better, the first application program can be recommended in the searching process, the probability that the high-quality application program is selected and used by a user is improved, and further the user experience is improved.

Fig. 1 illustrates an architecture diagram of a processing system for an application provided in an exemplary embodiment of the present application. As shown in fig. 1, the processing system for the application may include one or more terminal devices 101 and one or more servers 102. The number of the terminal devices and the servers in the processing system of the application program shown in fig. 1 is only an example, for example, the number of the terminal devices and the servers may be multiple, and the application does not limit the number of the terminal devices and the servers.

The terminal device 101 is a device used by a user, and includes at least one client, and the client includes at least one sub application (applet). For example, the client is a WeChat client that includes a News applet, a shopping applet, a gaming applet, and the like. The terminal apparatus 101 may include an AI model and a reference program set. Terminal device 101 may include, but is not limited to: smart phones (such as Android phones, iOS phones, etc.), tablet computers, portable personal computers, mobile internet devices (MID for short), and the like, which are not limited in the embodiments of the present invention. The server 102 refers to a background device for managing applications and providing services to the terminal device 101. The server 102 may include an AI model and a set of reference programs. The server 102 may include, but is not limited to, a cluster server.

In one embodiment, the process flow of the application may be performed by the server 102. Specifically, the method comprises the following steps: the server 102 performs quality identification on the application program deployed on the server through the processing method of the application program provided in the embodiment of the present application, so as to obtain a quality identification result of the application program. When the terminal device 101 requests the server 102 to acquire the application program, the server 102 provides the corresponding application program to the terminal device 101 according to the quality identification result of the application program; for example, the terminal device 101 requests to acquire an applet related to shopping, and the server 102 includes 10 applets related to shopping, but as can be seen from the quality recognition results of the applets, if only the quality of 3 applets related to shopping meets the push condition, the server 102 transmits the 3 applets related to shopping meeting the push condition to the terminal device 101.

In another embodiment, the processing flow of the application program may be executed by the terminal 101. The terminal device 101 identifies the quality of the application program in the terminal device according to the user behavior sequence stored in the terminal device or the sequence set provided by the server 102 by the application program processing method provided by the embodiment of the application, so as to obtain the quality identification result of the application program, and then recommends the corresponding application program for the user according to the quality identification result of the application program; for example: the terminal 101 includes 10 shopping-related applets, but according to the quality identification results of the applets, only the quality of 3 shopping-related applets meets the recommendation condition, and when a search instruction for a shopping-related applet sent by a user is received, the terminal 101 displays the 3 shopping-related applets meeting the recommendation condition to the user.

In the processing system of the application program shown in fig. 1, the processing flow of the application program mainly includes: the method comprises the steps that firstly, a terminal device 101 obtains main attributes (such as an identifier of a developer of a applet 1 in a chat platform, an identifier of an operator and a service address) of a first application and a reference program set (such as obtained through a server 102), wherein the reference program set comprises main attributes and quality characteristics (such as whether advertisement pushing is contained, the number of people concerned, the average daily use times and the like) of a plurality of reference applications, and secondly, a second application is obtained from the reference program set, wherein the second application can be one or more, and the second application is associated with the main attributes of the first application (such as the first application and the second application are developed by the same development company, or the development company of the first application is a subsidiary company of the development company of the second application). Third, the terminal device 101 predicts the quality characteristics of the first application program by using the quality characteristics of the second application program (if the first application program and the plurality of second application programs are developed by company a and each second application program contains advertisement push, it is predicted that the first application program also contains advertisement push), and fourth, the terminal device 101 outputs the quality recognition result of the first application program according to the quality characteristics of the first application program (if the quality score of the application program not containing the advertisement is set to 90, the quality score of the application program containing the advertisement is set to 70).

In the embodiment of the application, the terminal device or the server acquires the second application program from the reference program set according to the body attribute of the first application program, the body attribute of the second application program is associated with the body attribute of the first application program, the quality characteristic of the first application program is predicted by using the quality characteristic of the second application program, and the quality identification result of the first application program is output according to the quality characteristic of the first application program. Based on the idea of clustering, a second application program which is the same as the first application program is mined through the relevance between the main attributes, the quality characteristics of the first application program, which are obtained by adopting the quality characteristic prediction of the second application program, can reflect the overall quality of the first application program more accurately, and the more accurate quality identification result of the first application program is obtained based on the quality characteristics of the first application program.

In the processing system of the application program shown in fig. 1, the processing flow of another application program mainly includes: displaying a search page of a first client in the terminal device 101, wherein the first client comprises a first application (if the first application is the applet 1, the first client is the social platform, and the social platform comprises the applet 1), and obtaining, by the terminal device 101, a quality identification result of the first application obtained through the steps (i) to (iv). If the quality identification result of the first application program meets the recommendation condition (for example, the quality score is higher than 80 points), the information of the first application program (for example, the identification, the main body attribute, the category, the function brief introduction and the like of the first application program) is obtained through the memory of the terminal device or the server 102.

In the embodiment of the application, the terminal device displays a search page of the first client, obtains a quality identification result of the first application, the quality identification result of the first application is obtained by adopting the processing method of the application, and if the quality identification result of the first application meets the recommendation condition, information of the first application is obtained, and the information of the first application is displayed in the search page of the first client. Therefore, the application programs meeting the recommendation conditions are recommended, the probability that the high-quality application programs are selected and used by the user is improved, and the user experience is further improved.

Fig. 2 is a flowchart illustrating a processing method of an application according to an exemplary embodiment of the present application. The processing method of the application program may be executed by the terminal apparatus 101 or the server 102 shown in fig. 1; as shown in fig. 2, the processing method of the application program includes, but is not limited to, the following steps 201 to 204. The following describes in detail a processing method of an application provided in an embodiment of the present application, taking a terminal device as an example:

201. the terminal equipment acquires the main body attribute and the reference program set of the first application program.

The first application refers to any installation-free application in the terminal device, and specifically may refer to a sub-application (applet) in a client of the terminal device; for example, the first application is an applet in the IM client.

The subject attributes of the first application include, but are not limited to, an identifier of a developer of the first application, an identifier of an operator, a service address, and the like, which can distinguish the application. For example, the subject attributes of the application 1 include: the developer is a software development company A, and the service address is 165.28.3.16; the subject attributes of the application 2 include: the developer is software development company B, the operator is XX operation company, and the service address is 172.25.37.86.

The reference program set comprises subject attributes and quality characteristics of a plurality of reference applications, i.e. the quality characteristics of the reference applications in the reference program set are known. The reference application program refers to a sub-application program which is parasitized in the first client side and is except the first application program; alternatively, the reference application refers to a sub-application hosted within the second client. For example, the first application is applet 1 in the social platform, and the reference applications are applets 2-10 in the social platform; alternatively, the reference application is applet 1 to applet 10 in the shopping platform. The quality feature refers to a parameter to which the quality of an application is measured. For example, the quality characteristics include whether or not to include advertisement push, number of people interested, average number of times used per day, and the like.

In one embodiment, the reference program set is stored in a local storage space of the terminal device (e.g., a storage space built in or extended from the terminal device, or a memory of a client running in the terminal device). In another embodiment, the terminal device obtains the reference set via the server (e.g., the terminal device sends a obtaining request to the server, where the obtaining request is used to request obtaining the reference set).

202. The terminal device obtains the second application program from the reference program set.

Wherein the second application refers to an application associated with the subject property of the first application. Specifically, the first application and the second application have at least one same body attribute, or the at least one body attribute of the first application is related to the at least one body attribute of the second application, for example, the development company a of the application 1 is a subsidiary of the development company B of the application 2.

In one embodiment, the terminal device determines the subject attribute associated with the first application according to the subject attribute of the first application, screens out the reference applications having the subject attribute associated with the subject attribute of the first application from the reference program set according to the subject attribute of the first application, and determines the reference applications as the second application.

For example, assuming that the subject attribute of the first application is "company a development" and company B is a subsidiary of company a, the terminal device determines the associated subject attributes as "company a development" and "company B development" according to "company a development". Then, the terminal device screens out the reference applications 1 to 5 whose subject attributes are "company a development" or "company B development" from the reference program set, and determines the reference applications 1 to 5 as the second applications.

203. And the terminal equipment adopts the quality characteristics of the second application program to predict the quality characteristics of the first application program.

The quality characteristics refer to parameters which are referred to when the quality of the application program is measured; for example, the quality characteristics include whether or not to include advertisement push, number of people interested, average number of times used per day, and the like. Specifically, the terminal device obtains the quality characteristics of the second application program through the reference program set, and predicts the quality characteristics of the first application program through the quality characteristics of the second application program in the modes of data mining, deep learning, natural language processing and the like. For example, assuming that each second application carries a feature for displaying an advertisement in use, the terminal device predicts that the feature for displaying an advertisement in use is included in the quality features of the first application.

204. And the terminal equipment outputs the quality identification result of the first application program according to the quality characteristic of the first application program.

Wherein the quality identification result is used for indicating the quality of the first application program; for example, indicating that the first application is a premium application or a rogue application. In one implementation mode, the terminal device judges the quality of the first application program according to the quality characteristics of the first application program to obtain a quality identification result of the first application program; for example, the quality feature of the application 1 is "including false information", and the terminal device determines that the application 1 is a poor application (quality identification result) based on the "including false information".

Fig. 3 is a flowchart illustrating a processing method of an application according to an exemplary embodiment of the present application. The processing method of the application program may be executed by the terminal apparatus 101 or the server 102 shown in fig. 1; as shown in fig. 3, the processing method of the application program includes, but is not limited to, the following steps 301 to 306. The following describes in detail a processing method of an application provided in an embodiment of the present application, taking a terminal device as an example:

301. the terminal equipment acquires the main body attribute and the reference program set of the first application program.

The specific implementation of step 301 may refer to the implementation of step 201 in fig. 2, and is not described herein again.

302. And the terminal equipment constructs a main body relation graph.

The main body relation graph is used for establishing connection between the first application program node and the reference application program node through the main body attribute. The main body relation graph comprises attribute nodes and program nodes, the attribute nodes comprise main body attribute nodes of a first application program and main body attribute nodes of a reference application program, the program nodes comprise the first application program nodes and the reference application program nodes, and one attribute node is connected with a plurality of program nodes.

In one embodiment, the terminal device constructs a subject relationship graph from the first application and the set of reference routines. Specifically, taking the first application as an example: generating a program node of a first application program. Secondly, generating attribute nodes of the first application program, establishing connection between the program nodes of the first application program and the attribute nodes of the first application program, and enabling each attribute node to correspond to a main attribute of the first application program. And generating a program node of the second application program associated with the attribute node of the first application program, and establishing connection between the attribute node of the first application program and the program node of the second application program. And fourthly, continuing to establish the connection of other reference application program nodes in the reference program set in the mode of the reference step III until the program nodes in the main body relation graph contain all the reference application program nodes in the reference program set.

FIG. 4a illustrates a body relationship diagram provided by an exemplary embodiment of the present application. As shown in FIG. 4a, the subject relationship graph includes attribute node A through attribute node C and program node D through program node K. Assume that a first application corresponds to the program node G in fig. 4a, the first application has a body attribute 1 to a body attribute 3, respectively corresponds to the attribute node a to the attribute node C, and the program node D connected to the attribute node a, and the program node F and the program node I are program nodes of a second application having an association relationship with the body attribute 1 corresponding to the attribute node a. Similarly, the program node E and the program node H are program nodes of the second application program having an association relationship with the body attribute 2 corresponding to the attribute node B, and the program node J and the program node K are program nodes of the second application program having an association relationship with the body attribute 3 corresponding to the attribute node C. It should be noted that the number of attribute nodes, program nodes, and the structure of the main body relationship diagram in fig. 4a are only examples, and do not constitute practical limitations of the present application. For example, also connected to the property node a is a program node L, and also connected to the program node D is a property node M.

303. The terminal equipment splits the main body relational graph according to the attribute nodes to obtain a plurality of sub relational graphs, obtains a target sub relational graph from the plurality of sub relational graphs, and determines a second application program from the target sub relational graph.

The sub-relationship graph is used for determining a plurality of application programs which have an association relationship with the subject attribute (for example, the development company a of the application program 1 is a sub-company of the development company B of the application program 2, and the development companies of the application program 3 and the application program 4 are both the development company a). Each sub-relational graph comprises an attribute node and a plurality of program nodes. For example, if the sub-relationship fig. 1 includes the attribute node 1 and the program nodes 1 to 3, it indicates that all of the 3 application programs corresponding to the program nodes 1 to 3 have an association relationship with the body attribute corresponding to the attribute node 1. The target sub-relational graph is a sub-relational graph containing a first application program node in the program node. It is understood that the application corresponding to the program node other than the first application node in the target sub-relationship graph is the second application.

In one embodiment, the terminal device splits the main body relational graph according to the attribute nodes to obtain a plurality of sub-relational graphs (if the main body relational graph comprises N attribute nodes, and N is a positive integer, the main body relational graph can be split to obtain N sub-relational graphs), determines the target sub-relational graph, and determines the second application program from the target sub-relational graph. Specifically, each attribute node is taken as a central point, a program node connected with the attribute node is reserved, a sub-relationship graph including a first application program node (namely, the attribute node of the target sub-relationship graph is a main attribute node of the first application program) in the program node is determined as a target sub-relationship graph, and a reference application program corresponding to the program node except the first application program node in the target sub-relationship graph is determined as a second application program. It can be understood that by splitting the body relationship diagram, referring to a manner of obtaining the second application program associated with the first application program, the associated application programs of the respective reference application programs in the reference program set can be obtained, and by the method provided by the embodiment of the present application, the quality characteristics of the respective reference application programs in the reference program set can be updated.

Fig. 4b illustrates a sub-relationship diagram provided by an exemplary embodiment of the present application. The sub-relationship diagram shown in fig. 4B is obtained by splitting the attribute node B in fig. 4a, and includes the attribute node B, the program node G, the program node H, and the program node E, which have an association relationship with the subject attribute 2 corresponding to the attribute node B. Since the program node G in the sub-relationship graph is the first application node, the sub-relationship graph is the target sub-relationship graph. And the reference application program corresponding to the program node H and the program node E is a second application program of which the body attribute is associated with the body attribute of the first application program.

Optionally, if only the second application associated with the first application needs to be acquired, the target sub-relationship graph may be directly constructed through the body attribute of the first application and the reference program set, so as to determine the second application associated with the first application. Or screening the reference program set through the main body attribute of the first application program to obtain a second application program.

304. And the terminal equipment constructs a characteristic diagram of the first application program according to the quality characteristic of the second application program and the target sub-relational diagram.

The characteristic graph is obtained by attaching the main body attributes corresponding to the attribute nodes in the target sub-relational graph to the connecting edges between the program nodes. The program nodes in the feature graph comprise a first application program node and a second application program node, the second application program node carries the quality features of the second application program (the quality features of the second application program can be extracted through the historical use record of the second application program), and an edge formed by connecting any two program nodes carries the main attribute of the first application program.

In one embodiment, the terminal device connects the program nodes in the target sub-relational graph with each other, and deletes the attribute node in the target sub-relational graph to obtain the feature graph of the first application program. Fig. 4c illustrates a feature diagram provided by an exemplary embodiment of the present application. The characteristic diagram shown in fig. 4c is constructed according to fig. 4B, the program node E and the program node H carry the quality characteristics of the second application program, and the connection edge between the program nodes carries the main attribute 2 corresponding to the attribute node B.

305. And the terminal equipment calls the graph network model to carry out prediction processing on the characteristic graph to obtain the quality characteristic of the first application program.

The graph network model is used for predicting the quality characteristics of the first application program according to the characteristic graph. Graph Network models include, but are not limited to, Graph Convolutional neural networks (GCNs).

In one embodiment, the number of the body attributes of the first application is 1, that is, the number of the feature maps is 1, and the terminal device adopts the graph network model to fuse the features in the feature maps. Fig. 4d shows a flowchart for extracting quality features of an application in a feature map according to an exemplary embodiment of the present application. As shown in fig. 4d, the program node G is a first application program node, and the terminal device invokes the GCN to perform feature extraction on the feature map of the first application program, so as to obtain the quality feature of the first application program. For example, assuming that the program node E and the program node H both carry features for displaying an advertisement in use, and the subject attribute 2 is XX operation company, the quality features of the first application obtained by feature extraction of the feature map by the GCN include features for displaying an advertisement in use.

In another embodiment, the number of the body attributes of the first application is N, that is, the number of the feature maps is N, where N is a positive integer greater than 1. Fig. 4e shows a flowchart of another method for extracting quality features of an application program in a feature map according to an exemplary embodiment of the present application. As shown in fig. 4e, the program node G is a first application program node, and the terminal device invokes the GCN to perform feature extraction on each feature map respectively to obtain feature components of N first application programs, and then performs weighted aggregation on the feature components of the N first application programs through an Attention (Attention) mechanism, or performs aggregation on the feature components of the N first application programs through a concatenation (Concat) manner to obtain quality features of the first application programs.

The attention mechanism is used to increase the weight of the key feature components in the feature components of the first application, thereby increasing the influence of the key feature components of the first application on the quality features of the first application. For example, for a news applet, the key feature components are the number of users browsing, the number of praise, the authenticity of an article, and the like; for the shopping applet, the key feature components are the number of users purchased, quality evaluation of the product, and the like. Assuming that in fig. 4e, the weight of the feature component 1(f1) of the first application extracted by the GCN is 0.3, the weight of the feature component 2(f2) is 0.7, and the weight of the feature component 3(f3) is 0.4, the quality features of the first application obtained by weight aggregation are: f1 × 0.3+ f2 × 0.7+ f3 × 0.4; the quality characteristics of the first application program obtained by polymerization in a serial connection mode are as follows: f1+ f2+ f 3.

In another embodiment, the quality feature of the first application obtained by the terminal device is used as an intermediate quality feature, and the intermediate quality feature is subjected to stacking processing to obtain the quality feature of the first application. Fig. 4f shows a flowchart for extracting quality features of an application program in a feature map according to an exemplary embodiment of the present application. As shown in fig. 4f, the terminal device refers to the embodiment of obtaining the quality features of the first application program, obtains the quality features of each reference application program in the reference program set, constructs a new main body relationship diagram according to the intermediate quality features and the quality features of each reference application, further obtains an updated target sub-relationship diagram, obtains an updated feature diagram through the updated target sub-relationship diagram, and invokes the GCN to perform feature extraction on the updated feature diagram respectively, so as to obtain quality feature components of the first application program, and performs aggregation on the quality feature components of the first application program, so as to obtain the intermediate quality features. And repeating the steps to stack the medium quality features for M times to obtain the quality features of the first application program, wherein M is a positive integer. The neighborhood of the first application program can be enlarged through stacking processing, and more comprehensive and complete quality characteristics of the first application program are obtained, so that the accuracy of the quality identification result of the first application program is improved.

306. And the terminal equipment outputs the quality identification result of the first application program according to the quality characteristic of the first application program.

In one embodiment, the terminal device updates the reference set by adding the body attributes and quality features of the first application to the reference set. And then, identifying the quality characteristics of the first application program by adopting a classifier included in the graph network model to obtain a quality identification result of the first application program. Specifically, the classifier scores the quality of the first application program according to the score rule of each quality feature, and counts the scores of each quality feature of the first application program to obtain a quality identification result of the first application program. For example, the quality score of the applet whose user approval number is less than 1 ten thousand is set to 30, the quality score of the applet whose user approval number is between 1 ten thousand and 10 ten thousand is set to 50, the quality score of the applet whose user approval number is between 10 ten thousand and 50 ten thousand is set to 70, and the quality score of the applet whose user approval number is more than 50 ten thousand is set to 90. For another example, the quality score of the applet carrying the advertisement information is set to 60, and the quality score of the applet not carrying the advertisement information is set to 80. Assuming that the applet 1 has 3 quality features and the quality scores of the 3 quality features are 65, 90 and 70 respectively, the quality score in the quality recognition result of the applet 1 is 75.

Optionally, the quality features of the first application program are obtained by manual labeling or by performing feature extraction on the historical behavior record of the application program, and the obtained quality features are compared with the quality of the first application program obtained in the step 203, and the graph network model is trained (for example, parameters in the graph network model are adjusted) by using a gradient backhaul method, so that the optimized graph network model is obtained.

The following describes in detail a processing method of an application program provided in an embodiment of the present application by using an example: setting the quality of an application program of 'bus taking code scanning' required to be identified by the terminal equipment, wherein the 'bus taking code scanning' is developed by a development company XX and operated by an operation company YY; and the service address of the bus-taking code-scanning is a network address xxxxx; thus, the subject attributes of the "bus ride sweep code" include: development company XX, operations company YY and network address xxxxx. The processing method flow of the application program in this example is as follows:

firstly, a main body relation diagram of 'bus riding code sweeping' is constructed. As shown in fig. 4g, "bus taking code scanning" is taken as a program node, and 3 main attributes of the program node are taken as attribute nodes respectively; and establishing connection between a program node 'bus taking code scanning' and an attribute node 'development company XX', 'network address' and 'operation company YY', respectively. And then, sequentially acquiring the reference application programs from the reference program set as program nodes, acquiring the main attributes of the reference application programs as attribute nodes, and respectively connecting the program nodes and the attribute nodes according to attribute relevance. As shown in fig. 4g, the attribute node "development company XX" is further connected to program nodes "bus card two-dimensional code riding" and "bus e-pass riding code", which means that the program nodes of the reference application programs and the "bus riding scanning code" have the same or related subject attributes (i.e. the development company XX), and further means that the application programs of the reference application programs and the "bus riding scanning code" are developed by the development company XX or related companies (e.g. subsidiaries of the development company XX). Similarly, the attribute node "network address xxx" is further connected with program nodes "public transport e-pass bus taking code", "AA subway bus trip" and "bus taking code scanning assistant"; the reference applications and the "bus scan code" application are all served by a server with a network address xxxxx or a server associated with the xxxxx server (as with other servers in a local area network). Similarly, the attribute node "operation company YY" is also connected with the program node "riding" and "bus code sweeping riding"; the reference applications and the "bus scan code" are all operated by the operating company YY or its related companies (e.g., the subsidiary companies of the operating company YY).

Secondly, the main body relational graph is split to obtain a plurality of sub-relational graphs. In this example, referring to fig. 4g, each circular dotted box includes an attribute node and a plurality of program nodes connected to the attribute node; then, the main body relational graph can be split according to a circular dashed frame, and three target sub-relational graphs related to the application program of the bus scanning code can be obtained. And then other reference application programs in the sub-relation graph, such as 'AA subway bus trip', 'bus e-pass bus code', 'bus taking' and the like can be determined as a second application program.

And then, generating a characteristic graph according to the second application program and the target sub-relational graph. Referring to fig. 4h, the feature graph includes a development company sub-graph, a network address sub-graph, and an operation company sub-graph. The meaning of each node in fig. 4h is shown in table 1 below:

table 1: node meaning table

Deleting the connection between the attribute node A and the program nodes H and E, establishing the connection between the program nodes (such as establishing the connection between the program node G and the program node E and establishing the connection between the program node E and the program node H), and adding the development company XX into the connection among the program nodes, so that the connection among the program nodes carries the information of the development company XX, thereby obtaining the characteristic graph corresponding to the development company subgraph. And similarly, respectively generating a feature graph of the network address subgraph and the operating company subgraph according to the second application program and the target sub-relationship graph.

And then, performing feature extraction on each feature map to obtain a quality feature component, and performing aggregation processing on the quality feature components to obtain the quality feature of the first application program. As shown in fig. 4h, calling a graph convolutional neural network model (such as GCN) to perform feature extraction processing on the three feature graphs respectively to obtain three quality feature components of the "bus taking code scanning"; and performing weighted aggregation processing on the three quality characteristic components by adopting an attention mechanism to obtain the intermediate quality characteristic of the bus scanning code.

And finally, superposing the intermediate quality characteristics of the bus scanning codes to obtain the quality characteristics of the bus scanning codes. As shown in fig. 4i, the quality characteristics of the second application programs such as "AA subway bus trip", "bus e-pass bus code", "bus taken", and the like can be obtained in the above manner, and are calculated once or multiple times in the above manner to obtain the quality characteristics (i.e., stacking processing) of "bus taking code scanning", and the final result after stacking is classified by the classifier to obtain the quality recognition result of "bus taking code scanning".

In the embodiment of the application, a main body relational graph is constructed according to main body attributes of a first application program, namely a reference program set, the main body relational graph is split to obtain a plurality of sub-relational graphs, a target sub-relational graph is determined, a second application program is further determined, a characteristic graph of the first application program is constructed according to the second application program and the target sub-relational graph, the characteristic graph is predicted through a graph network model to obtain quality characteristics of the first application program, and a quality identification result of the first application program is output according to the quality characteristics of the first application program. Based on the idea of clustering, a second application program which is the same as the first application program is mined through the relevance between the main attributes, the quality characteristics of the first application program, which are obtained by adopting the quality characteristic prediction of the second application program, can reflect the overall quality of the first application program more accurately, and the more accurate quality identification result of the first application program is obtained based on the quality characteristics of the first application program.

Fig. 5 is a flowchart illustrating a processing method of another application according to an exemplary embodiment of the present application. The processing method of the application program may be executed by the terminal apparatus 101 shown in fig. 1; as shown in fig. 5, the processing method of the application program includes, but is not limited to, the following steps 501 to 504. Wherein:

501. the terminal device displays a search page of the first client.

The first client is any client on the terminal device, the first client comprises a first application program, and the first application program refers to any sub-application program parasitic in the first client. For example, a mobile phone includes a social APP (i.e., a client), and the social APP includes applets 1 to 10. For another example, the computer includes shopping software 1, and shopping software 1 includes sub-applications 1 to 3.

502. The terminal equipment acquires a quality identification result of the first application program.

In an embodiment, the terminal device obtains a quality identification result of the first application program from the local storage space, where the identification result of the first application program is obtained by the terminal device using the embodiment described in the embodiment of fig. 2, and the quality identification result of the first application program includes a quality score of the first application program. For example, the quality recognition result of the application 1 is a shopping program, and the quality score is 80.

In another embodiment, the terminal device obtains the quality identification result of the first application program from the server, and the identification result of the first application program is obtained by the server by using the embodiment described in the embodiment of fig. 2.

503. And if the quality identification result of the first application program meets the recommendation condition, the terminal equipment acquires the information of the first application program.

Wherein, the condition of meeting the recommendation means that the quality score of the first application is greater than the quality threshold. For example, if the quality threshold is 70 points, the quality score of the application 1 is 80 points, and the quality score of the application 2 is 60 points, it is determined that the application 1 satisfies the recommendation condition and the application 2 does not satisfy the recommendation condition. The information of the first application includes at least one of: identification of the first application, subject attributes, categories, function profiles.

In one embodiment, if the terminal device determines that the first application satisfies the recommendation condition, the terminal device obtains information of the first application from the server or the local storage space, and adds the information of the first application to the recommendation list. And the terminal equipment sorts the application programs in the recommendation list according to the sequence of the quality from high to low.

In another embodiment, if the quality score of the first application is smaller than the penalty threshold, the terminal device performs penalty processing on the first application. And the penalty threshold is smaller than the quality threshold, namely, the application programs meeting the recommendation condition cannot be penalized, and the penalized application programs cannot be recommended. The penalty processing includes: the method comprises the steps that the terminal device conducts shielding processing on a first application program (namely the terminal device does not display when a user searches the first application program), deleting processing (namely the terminal device deletes the first application program), sending modification prompt information to an operator of the first application program, wherein the modification prompt information comprises possible problems of the first application program (for example, false information exists in an article issued by the operator is prompted), and the modification prompt information is used for prompting the operator to conduct self-checking and improvement on the possible problems.

504. The terminal device displays information of the first application program in a search page of the first client.

In one embodiment, the search page of the first client includes a recommendation bar, and the terminal device displays information in the recommendation list in the recommendation bar of the first client. Fig. 6a illustrates a search page diagram of a first client according to an exemplary embodiment of the present application. As shown in fig. 6a, 601 is a search bar of the first client, and when a user needs to search an applet, the user clicks the search bar, inputs a keyword of the applet that needs to be searched, and clicks a "search for one" button to perform a search. 602 is a recommendation column of the first client, the terminal device displays information in a recommendation list in 602, and the applets in the recommendation list are the applets whose quality identification results meet recommendation conditions. The arrangement sequence of the small programs in the recommendation list is obtained by sequencing the quality scores of the small programs from high to low; or after the relevance between the small programs and the historical search records of the user, the quality scores of the small programs and other factors are subjected to comprehensive scores, the small programs are sorted from high to low according to the comprehensive scores.

In another embodiment, after receiving a search instruction of a user, the terminal performs comprehensive scoring according to multiple factors such as the quality scores of the application programs and the like according to the relevance degrees of the keywords, sorts the application programs corresponding to the search instruction according to the sequence of the comprehensive scores from high to low, and displays the sorted result in a search result column. The specific implementation manner may refer to an implementation manner in which information in the recommendation list is displayed in the recommendation bar, and details are not described herein. Fig. 6b illustrates a search result page diagram of a first client according to an exemplary embodiment of the present application. As shown in FIG. 6b, the user enters "shopping" at 601 and clicks the "search for" button. 603 is a search result column of the first client, the terminal device determines a corresponding application according to "shopping", sorts information of the searched application, and displays the sorted result in the search result column 603.

Fig. 7 is a flowchart illustrating a processing method of another application according to an exemplary embodiment of the present application. The processing method of the application program can be executed by the terminal device 101 shown in fig. 1, or by the terminal device 101 shown in fig. 1 and the server 102 in an interactive manner; as shown in fig. 7, the processing method of the application program includes, but is not limited to, the following steps 701 to 708. Wherein:

701. the terminal equipment acquires the main body attribute and the reference program set of the first application program.

702. The terminal device obtains the second application program from the reference program set.

703. And the terminal equipment adopts the quality characteristics of the second application program to predict the quality characteristics of the first application program.

704. And the terminal equipment outputs the quality identification result of the first application program according to the quality characteristic of the first application program.

The specific implementation of steps 701 to 704 can refer to the implementation of steps 201 to 204 in fig. 2, and will not be described herein again. Steps 701 to 704 may be executed by the server.

705. The terminal device displays a search page of the first client.

706. The terminal equipment acquires a quality identification result of the first application program.

707. And if the quality identification result of the first application program meets the recommendation condition, the terminal equipment acquires the information of the first application program.

708. The terminal device displays information of the first application program in a search page of the first client.

The specific implementation of step 705 to step 708 can refer to the implementation of step 501 to step 504 in fig. 5, and will not be described herein again.

The following describes in detail the processing method of the application program provided in the embodiment of the present application by using a complete example: let 1 be a newly developed applet for company a and responsible for operational maintenance by company B, and let 1 be an applet in APP 1. The terminal device acquires the subject attributes of the applet 1, and the subject attributes of the applet 1 include a development company (i.e., company a) and an operation company (i.e., company B). The terminal device acquires a reference program set including the subject attributes and quality characteristics of the applets 2 to 100. The terminal device screens out, from the reference program set, applets 13 and 24 developed by company a, applets 5 and 31 developed by company C (company C is a subsidiary of company a), and applets 12, applets 63 and applets 78 taken care of operation and maintenance by company B. Assuming that the applet 5, the applet 13, the applet 24 and the applet 31 are all shopping applets and that none of the applet 12, the applet 63 and the applet 78 contain advertisement information, the terminal device predicts the applet 1 as a shopping applet containing no advertisement information based on the characteristics of these 7 applets. The terminal device then obtains the quality recognition result of the applet 1 for the shopping applet (quality feature) not containing advertisement information according to the applet 1.

When the user opens the search page of APP1, the terminal device acquires the quality recognition results of other applets in applet 1 and APP 1. Firstly, based on the quality scores contained in the quality identification results, selecting the small programs with the quality scores higher than the quality threshold value and the small programs with the quality scores lower than the penalty threshold value from the small programs. Then, the small programs with the quality scores higher than the quality threshold are sorted in the order of the quality scores from high to low, and the information of the small programs is sequentially displayed in a recommendation column of the search page according to the sorting result. As shown in fig. 6a, the terminal device displays the 3 applets with the highest quality scores among the applets in the recommendation field 602. Wherein, the quality score of "XX small program" < "> the quality score of" YY small program > < "> the quality score of ZZ small program > the quality threshold. Similarly, as shown in fig. 6b, after the search keyword is acquired, the terminal device matches the corresponding applets according to the search keyword, sorts the applets whose quality scores are higher than the quality threshold value in the order of the quality scores from high to low, and sequentially displays the information of the applets in the search result column 603 of the search page according to the sorting result. And then, carrying out punishment processing such as shielding and deleting the small program with the quality score lower than the punishment threshold value or sending modification prompt information to the operator of the small program.

In the embodiment of the application, the terminal device acquires the second application program from the reference program set according to the main body attribute of the first application program, the main body attribute of the second application program is associated with the main body attribute of the first application program, the quality characteristic of the first application program is predicted by adopting the quality characteristic of the second application program, and the quality identification result of the first application program is output according to the quality characteristic of the first application program. Based on the idea of clustering, a second application program which is the same as the first application program is mined through the relevance between the main attributes, the quality characteristics of the first application program, which are obtained through the quality characteristic prediction of the second application program, can reflect the overall quality of the first application program more accurately, and the quality identification technology is simple, efficient, low in cost consumption and higher in practicability through obtaining the more accurate quality identification result of the first application program based on the quality characteristics of the first application program; when the first client side searches the application programs, if the quality identification result of the first application program shows that the quality of the first application program is better, the first application program can be recommended in the searching process, the probability that the high-quality application program is selected and used by a user is improved, and further the user experience is improved.

While the method of the embodiments of the present application has been described in detail above, to facilitate better implementation of the above-described aspects of the embodiments of the present application, the apparatus of the embodiments of the present application is provided below accordingly.

Referring to fig. 8, fig. 8 is a schematic structural diagram of a processing apparatus of an application according to an exemplary embodiment of the present application, where the processing apparatus of the application may be mounted on a terminal device or a server in the foregoing method embodiment. The processing means of the application shown in fig. 8 may be adapted to perform some or all of the functions of the method embodiments described above with reference to fig. 2, 3 and 7. Wherein, the detailed description of each unit is as follows:

an obtaining unit 801, configured to obtain a body attribute and a reference program set of a first application program, where the reference program set includes the body attribute and quality characteristics of multiple reference application programs; and means for obtaining a second application from the set of reference applications, the second application being at least one reference application in the set of reference applications, and a subject attribute of the second application being associated with a subject attribute of the first application;

the processing unit 802 is configured to predict the quality characteristic of the first application program by using the quality characteristic of the second application program, and output a quality identification result of the first application program according to the quality characteristic of the first application program.

In an embodiment, the processing unit 802 is further configured to obtain a second application program from the reference program set, and specifically configured to:

constructing a main body relational graph, wherein the main body relational graph comprises attribute nodes and program nodes, the attribute nodes comprise main body attribute nodes of the first application program and main body attribute nodes of the reference application program, the program nodes comprise first application program nodes and reference application program nodes, and one attribute node is connected with a plurality of program nodes;

acquiring a target sub-relational graph from the plurality of sub-relational graphs, and determining a second application program from the target sub-relational graph; the attribute nodes of the target sub-relational graph are subject attribute nodes of the first application program, and the program nodes of the target sub-relational graph comprise the first application program nodes and second application program nodes.

In an embodiment, the processing unit 802 is further configured to predict the quality characteristic of the first application by using the quality characteristic of the second application, and specifically configured to:

constructing a feature graph of the first application program according to the quality features of the second application program and the target sub-relational graph;

and calling a graph network model to perform prediction processing on the characteristic graph to obtain the quality characteristic of the first application program.

In an embodiment, the processing unit 802 is further configured to construct a feature map of the first application according to the quality feature of the second application and the target sub-relationship map, and specifically configured to:

the program nodes in the feature graph comprise the first application program node and the second application program node, the second application program node carries the quality feature of the second application program, and an edge formed by connecting any two program nodes carries the main attribute of the first application program.

In one embodiment, the graph network model comprises a graph convolution network;

the processing unit 802 is further configured to call a graph network model to perform prediction processing on the feature graph to obtain a quality feature of the first application program, and specifically configured to:

and calling the graph convolution network to extract the quality characteristics of the first application program from the characteristic graph.

calling N layers of graph convolution layers to respectively extract N quality characteristic components of the first application program from N characteristic graphs; wherein, a layer of graph convolution layer is used for carrying out feature extraction processing on a feature graph to obtain a quality feature component of the first application program;

calling the graph convolution network to carry out convolution processing on the feature graph to obtain an intermediate quality feature;

and stacking the intermediate quality features to obtain the quality features of the first application program.

In one embodiment, the graph network model includes a classifier;

the processing unit 802 is further configured to output a quality identification result of the first application according to the quality characteristic of the first application, and specifically configured to:

and calling the classifier to classify the quality characteristics of the first application program to obtain a quality identification result of the first application program.

In an embodiment, the processing unit 802 is further configured to:

adding the subject attributes and quality features of the first application to the reference program set;

the first application and the reference application are both installation-free applications; the first application program refers to any sub-application program parasitic in the first client side; the reference application program is a sub application program which is parasitized in the first client except the first application program; alternatively, the reference application is a sub-application hosted in the second client.

According to an embodiment of the present application, some steps involved in the processing method of the application shown in fig. 2, fig. 3 and fig. 7 may be executed by each unit in the processing device of the application shown in fig. 8. For example, step 201 and step 202 shown in fig. 2 may be performed by the acquisition unit 801 shown in fig. 8, and step 203 and step 204 may be performed by the processing unit 802 shown in fig. 8. Step 301 shown in fig. 3 may be performed by the acquisition unit 801 shown in fig. 8, and steps 302 to 306 may be performed by the processing unit 802 shown in fig. 8.

Steps

701 and 702 shown in fig. 7 may be performed by the acquisition unit 801 shown in fig. 8, and steps 703 and 704 may be performed by the processing unit 802 shown in fig. 8. The units in the processing device of the application program shown in fig. 8 may be respectively or entirely combined into one or several other units to form one or several other units, or some unit(s) may be further split into multiple functionally smaller units to form one or several other units, which may achieve the same operation without affecting the achievement of the technical effect of the embodiments of the present application. The units are divided based on logic functions, and in practical application, the functions of one unit can be realized by a plurality of units, or the functions of a plurality of units can be realized by one unit. In other embodiments of the present application, the processing device of the application program may also include other units, and in practical applications, the functions may also be implemented by being assisted by other units, and may be implemented by cooperation of a plurality of units.

According to another embodiment of the present application, a processing apparatus of an application program as shown in fig. 8 may be constructed by running a computer program (including program codes) capable of executing steps involved in the respective methods as shown in fig. 2, 3 and 7 on a general-purpose computing apparatus such as a computer including a processing element such as a Central Processing Unit (CPU), a random access storage medium (RAM), a read-only storage medium (ROM) and a storage element, and a processing method of an application program of an embodiment of the present application may be implemented. The computer program may be recorded on a computer-readable recording medium, for example, and loaded and executed in the above-described computing apparatus via the computer-readable recording medium.

Based on the same inventive concept, the principle and the advantageous effect of the problem solving of the processing apparatus for the application program provided in the embodiment of the present application are similar to the principle and the advantageous effect of the problem solving of the processing method for the application program in the embodiment of the present application, and for brevity, the principle and the advantageous effect of the implementation of the method may be referred to, and are not described herein again.

Referring to fig. 9, fig. 9 is a schematic structural diagram of a processing apparatus of another application program according to an exemplary embodiment of the present application, where the processing apparatus of the application program may be mounted on a terminal device in the foregoing method embodiment. The processing means of the application shown in fig. 9 may be adapted to perform some or all of the functions in the method embodiments described above with reference to fig. 5 and 7. Wherein, the detailed description of each unit is as follows:

a display unit 901, configured to display a search page of a first client, where the first client includes a first application program, and the first application program refers to any sub-application program hosted in the first client;

a processing unit 902, configured to obtain a quality identification result of the first application, where the quality identification result of the first application is obtained by using the method in the method embodiment described in the foregoing fig. 2; if the quality identification result of the first application program meets the recommendation condition, acquiring the information of the first application program;

the display unit 901 is further configured to display information of the first application program in a search page of the first client.

In one embodiment, the quality identification result comprises a quality score; the quality identification result of the first application program meets the recommendation condition, namely the quality score of the first application program is larger than a quality threshold; the information of the first application includes at least one of: identification, subject attributes, categories, function profiles of the first application;

the processing unit 902 is further configured to display information of the first application program in a search page of the first client, specifically to:

adding the information of the first application program into a recommendation list of the first client, wherein the recommendation list comprises a plurality of pieces of information to be recommended, and the information is sorted according to the sequence of the quality scores of the corresponding application programs from high to low;

displaying information in a recommendation list of the first client in a search page of the first client;

the processing unit 902 is further configured to:

if the quality score of the first application program is smaller than a penalty threshold, performing penalty processing on the first application program, wherein the penalty threshold is smaller than the quality threshold, and the penalty processing comprises at least one of the following steps: and shielding, deleting and sending the rectification prompt information to the operator of the first application program.

According to an embodiment of the present application, some steps involved in the processing method of the application program shown in fig. 5 and 7 may be executed by each unit in the processing apparatus of the application program shown in fig. 9. For example, step 501 and step 504 shown in fig. 5 may be performed by the display unit 901 shown in fig. 9, and step 502 and step 503 may be performed by the processing unit 902 shown in fig. 9. Step 705 and step 708 shown in fig. 7 may be performed by the display unit 901 shown in fig. 9, and step 706 and step 707 may be performed by the processing unit 902 shown in fig. 9. The units in the processing device of the application program shown in fig. 9 may be respectively or entirely combined into one or several other units to form one or several other units, or some unit(s) may be further split into multiple functionally smaller units to form one or several other units, which may achieve the same operation without affecting the achievement of the technical effect of the embodiments of the present application. The units are divided based on logic functions, and in practical application, the functions of one unit can be realized by a plurality of units, or the functions of a plurality of units can be realized by one unit. In other embodiments of the present application, the processing device of the application program may also include other units, and in practical applications, the functions may also be implemented by being assisted by other units, and may be implemented by cooperation of a plurality of units.

According to another embodiment of the present application, a processing apparatus of an application program as shown in fig. 9 may be constructed by running a computer program (including program codes) capable of executing steps involved in the respective methods as shown in fig. 5 and 7 on a general-purpose computing apparatus such as a computer including a processing element such as a Central Processing Unit (CPU), a random access storage medium (RAM), a read-only storage medium (ROM), and a storage element, and a processing method of an application program of an embodiment of the present application may be implemented. The computer program may be recorded on a computer-readable recording medium, for example, and loaded and executed in the above-described computing apparatus via the computer-readable recording medium.

Referring to fig. 10, fig. 10 is a schematic structural diagram illustrating a processing device for an application according to an exemplary embodiment of the present application, where the processing device for the application at least includes a processor 1001, a communication interface 1002, and a memory 1003. The processor 1001, the communication interface 1002, and the memory 1003 may be connected by a bus or in other manners, and in the embodiment of the present application, the connection by the bus is taken as an example. The processor 1001 (or Central Processing Unit (CPU)) is a computing core and a control core of the processing device of the application program, and can analyze various instructions in the terminal device and various data of the processing terminal device, for example: the CPU can be used for analyzing a power-on and power-off instruction sent to the terminal equipment by a user and controlling the terminal equipment to carry out power-on and power-off operation; the following steps are repeated: the CPU may transmit various types of interactive data between the internal structures of the terminal device, and so on. The communication interface 1002 may optionally include a standard wired interface, a wireless interface (e.g., WI-FI, mobile communication interface, etc.), which may be controlled by the processor 1001 for transceiving data; the communication interface 1002 can also be used for transmission and interaction of data inside the terminal device. The Memory 1003(Memory) is a Memory device in the terminal device, and stores programs and data. It is understood that the memory 1003 herein may include a built-in memory of the terminal device, and may also include an expansion memory supported by the terminal device. The memory 1003 provides storage space that stores the operating system of the terminal device, which may include, but is not limited to: android system, iOS system, Windows Phone system, etc., which are not limited in this application.

In one embodiment, the processing device of the application may refer to a terminal device or a server, such as the terminal device 101 or the server 102 shown in fig. 1. In this case, the processor 1001 performs the following operations by executing the executable program code in the memory 1003:

acquiring a main body attribute and a reference program set of a first application program through a communication interface 1002, wherein the reference program set comprises the main body attribute and quality characteristics of a plurality of reference application programs;

obtaining a second application from the set of references, the second application being at least one reference application in the set of references, and a subject attribute of the second application being associated with a subject attribute of the first application;

predicting a quality characteristic of the first application program by using a quality characteristic of the second application program;

As an alternative implementation, the specific implementation manner of the processor 1001 acquiring the second application program from the reference program set is as follows:

As an optional implementation manner, the specific implementation manner of the processor 1001 predicting the quality feature of the first application by using the quality feature of the second application is as follows:

As an optional implementation manner, the specific implementation manner of the processor 1001, according to the quality feature of the second application and the target sub-relationship graph, constructing the feature graph of the first application is as follows:

As an optional implementation, the graph network model includes a graph convolution network;

the specific implementation manner of the processor 1001 calling the graph network model to perform prediction processing on the feature graph to obtain the quality feature of the first application program is as follows:

As an optional implementation manner, the number of the main attributes of the first application is N, the number of the feature maps of the first application is N, the graph network model includes a graph convolution network, the graph convolution network includes N layers of graph convolution layers, and N is an integer greater than 1;

As an alternative embodiment, the graph network model includes a classifier;

the specific implementation manner of the processor 1001 outputting the quality identification result of the first application according to the quality feature of the first application is as follows:

As an alternative embodiment, the processor 1001, by running the executable program code in the memory 1003, further performs the following operations:

In another embodiment, the processing device of the application may refer to a terminal device, such as terminal device 101 shown in fig. 1. In this case, the processor 1101 performs the following operations by executing the executable program code in the memory 1103:

acquiring a quality identification result of the first application program through a communication interface 1102, wherein the quality identification result of the first application program is obtained by adopting the method in the method embodiment described in the above fig. 2;

if the quality identification result of the first application program meets the recommendation condition, acquiring the information of the first application program;

and displaying the information of the first application program in a search page of the first client.

As an optional implementation, the quality identification result includes a quality score; the quality identification result of the first application program meets the recommendation condition, namely the quality score of the first application program is larger than a quality threshold; the information of the first application includes at least one of: identification, subject attributes, categories, function profiles of the first application;

the specific implementation manner of the processor 1101 displaying the information of the first application program in the search page of the first client is as follows:

the processor 1101, by executing the executable program code in the memory 1103, also performs the following operations:

Based on the same inventive concept, the principle and the advantageous effect of the problem solving of the processing device of the application program provided in the embodiment of the present application are similar to the principle and the advantageous effect of the problem solving of the processing method of the application program in the embodiment of the present application, and for brevity, the principle and the advantageous effect of the implementation of the method may be referred to, and are not described herein again.

The embodiment of the present application further provides a computer-readable storage medium, where one or more instructions are stored in the computer-readable storage medium, and the one or more instructions are adapted to be loaded by a processor and to execute the processing method of the application program according to the above method embodiment.

Embodiments of the present application further provide a computer program product containing instructions, which when run on a computer, cause the computer to execute the processing method of the application program described in the above method embodiments.

It should be noted that, for simplicity of description, the above-mentioned embodiments of the method are described as a series of acts or combinations, but those skilled in the art should understand that the present application is not limited by the order of acts described, as some steps may be performed in other orders or simultaneously according to the present application. Further, those skilled in the art should also appreciate that the embodiments described in the specification are preferred embodiments and that the acts and modules referred to are not necessarily required in this application.

The steps in the method of the embodiment of the application can be sequentially adjusted, combined and deleted according to actual needs.

The modules in the device can be merged, divided and deleted according to actual needs.

Those skilled in the art will appreciate that all or part of the steps in the methods of the above embodiments may be implemented by hardware related to instructions of a program, which may be stored in a computer readable storage medium, which may include: flash disks, Read-Only memories (ROMs), Random Access Memories (RAMs), magnetic or optical disks, and the like.

While the invention has been described with reference to a preferred embodiment, it will be understood by those skilled in the art that various changes in form and detail may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.

Claims

1. A method for processing an application program, the method comprising:

acquiring a main body attribute and a reference program set of a first application program, wherein the reference program set comprises the main body attribute and quality characteristics of a plurality of reference application programs;

2. The method of claim 1, wherein said obtaining a second application from the reference program set comprises:

3. The method of claim 2, wherein predicting the quality characteristic of the first application using the quality characteristic of the second application comprises:

4. The method of claim 3, wherein constructing the feature map of the first application from the quality features of the second application and the target sub-relationship map comprises:

5. The method of claim 3, wherein the graph network model comprises a graph convolution network; the calling graph network model carries out prediction processing on the feature graph to obtain the quality feature of the first application program, and the method comprises the following steps:

6. The method of claim 3, wherein the number of subject attributes of the first application is N, the number of feature maps of the first application is N, the graph network model comprises a graph convolution network comprising N layers of graph convolution layers, N being an integer greater than 1;

the calling graph network model carries out prediction processing on the feature graph to obtain the quality feature of the first application program, and the method comprises the following steps:

7. The method of claim 3, wherein the graph network model comprises a graph convolution network; the calling graph network model carries out prediction processing on the feature graph to obtain the quality feature of the first application program, and the method further comprises the following steps:

8. The method of claim 3, wherein the graph network model comprises a classifier; the outputting the quality identification result of the first application program according to the quality characteristic of the first application program comprises:

9. The method of claim 1, further comprising:

10. A method for processing an application program, the method comprising:

acquiring a quality identification result of the first application program, wherein the quality identification result of the first application program is obtained by adopting the method of any one of claims 1 to 9;

11. The method of claim 10, wherein the quality identification result comprises a quality score; the quality identification result of the first application program meets the recommendation condition, namely the quality score of the first application program is larger than a quality threshold; the information of the first application includes at least one of: identification, subject attributes, categories, function profiles of the first application;

the displaying the information of the first application program in the search page of the first client includes:

the method further comprises the following steps:

12. An apparatus for processing an application program, comprising:

the device comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring the main body attribute and a reference program set of a first application program, and the reference program set comprises the main body attribute and the quality characteristic of a plurality of reference application programs; and means for obtaining a second application from the set of reference applications, the second application being at least one reference application in the set of reference applications, and a subject attribute of the second application being associated with a subject attribute of the first application;

and the processing unit is used for predicting the quality characteristics of the first application program by adopting the quality characteristics of the second application program and outputting the quality identification result of the first application program according to the quality characteristics of the first application program.

13. An apparatus for processing an application program, comprising:

a processing unit, configured to obtain a quality identification result of the first application, where the quality identification result of the first application is identified by using the method according to any one of claims 1 to 9; if the quality identification result of the first application program meets the recommendation condition, acquiring the information of the first application program;

the display unit is further configured to display information of the first application program in a search page of the first client.

14. An apparatus for processing an application program, comprising:

a memory storing computer readable instructions;

a processor coupled to the memory, the processor configured to execute the computer readable instructions to implement the processing method of the application program of any one of claims 1-9, or to implement the processing method of the application program of claim 10 or 11.

15. A computer-readable storage medium, characterized in that it stores one or more instructions adapted to be loaded by said processor and to carry out the processing method of an application according to any one of claims 1 to 9, or to implement the processing method of an application according to claim 10 or 11.