Installation package compliance detection method, installation package compliance detection device and readable medium

Abstract

The invention provides an installation package compliance detection method, which comprises the following steps of: detecting whether an image is contained in an installation package in an installation process of the installation package, and if the image is contained, analyzing the image to obtain first description information of the image, namely text information of a human language representing relevance among elements in the image; determining the compliance of the installation package according to the first description information; according to the embodiment of the invention, whether the image in the installation package meets the specification or not can be automatically analyzed in the installation process of the installation package, and the compliance detection opportunity of the application is advanced, so that a user can be prompted and alarmed in advance, and potential safety hazards caused by non-compliance applications are reduced. The invention further provides an installation package compliance detection device and a readable medium.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and in particular to an installation package compliance detection method, an installation package compliance detection device, and a readable medium. Background Art

[0002] Crimes committed through smart terminals are common, such as financial fraud, pornography, gambling, and drug-related crimes, which seriously endanger social security. Giving smart terminal devices the means to identify the above sensitive information can effectively remind users, improve their relevant awareness, and reduce the harm and property losses caused by sensitive information.

[0003] In the related technology, the compliance of applications (APPs) can be detected by voice recognition and text recognition. However, this technology has the following disadvantages: voice recognition and text recognition are both detected and judged during the use of the application, and the timing of compliance detection is delayed, which still poses security risks; moreover, the compliance detection is performed on the single identified keyword, and the accuracy is not high. Summary of the invention

[0004] The present disclosure provides an installation package compliance detection method, an installation package compliance detection device, and a readable medium.

[0005] In a first aspect, an embodiment of the present disclosure provides an installation package compliance detection method, comprising:

[0006] When it is detected during the installation process of the installation package that the installation package contains an image, the image is analyzed to obtain first description information of the image, where the first description information is text information in human language that represents the association between elements in the image;

[0007] The compliance of the installation package is determined according to the first description information.

[0008] On the other hand, an embodiment of the present disclosure also provides an installation package compliance detection device, comprising: one or more processors; a storage device, on which one or more programs are stored; when the one or more programs are executed by the one or more processors, the one or more processors implement the installation package compliance detection method as described above; one or more I / O interfaces, connected between the processor and the storage device, configured to implement information interaction between the processor and the storage device.

[0009] On the other hand, an embodiment of the present disclosure further provides a computer-readable medium having a computer program stored thereon, wherein the program, when executed, implements the installation package compliance detection method as described above.

[0010] The installation package compliance detection method provided by the embodiment of the present disclosure detects whether the installation package contains an image during the installation process of the installation package. If the image is contained, the image is analyzed to obtain the first description information of the image, that is, the text information in human language that characterizes the correlation between the elements in the image, and the compliance of the installation package is determined based on the first description information; the embodiment of the present disclosure can automatically analyze whether the image in the installation package complies with the regulations during the installation process of the installation package, and advance the timing of the compliance detection of the application, so that the user can be prompted with an alarm in advance to reduce the security risks caused by non-compliant applications. Moreover, the first description information can include phrases that can reflect information such as behavioral actions. Therefore, compliance detection based on the first description information can improve the accuracy of detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 Schematic diagram of the installation package compliance detection process provided by the embodiment of the present disclosure Figure 1 ;

[0012] Figure 2 A schematic diagram of a process for generating first description information of an image provided by an embodiment of the present disclosure;

[0013] Figure 3 A schematic diagram of a process for detecting compliance of an installation package provided in an embodiment of the present disclosure;

[0014] Figure 4 A schematic diagram of a process for sending the first alarm information provided in an embodiment of the present disclosure;

[0015] Figure 5 A schematic diagram of a process for recognizing text information in an image provided by an embodiment of the present disclosure;

[0016] Figure 6 Schematic diagram of the installation package compliance detection process provided by the embodiment of the present disclosure Figure 2 ;

[0017] Figure 7 A schematic diagram of the structure of an installation package compliance detection device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0018] Example embodiments will be described more fully below with reference to the accompanying drawings, but the example embodiments may be embodied in different forms and should not be construed as limited to the embodiments set forth herein. On the contrary, the purpose of providing these embodiments is to make the present disclosure thorough and complete and to enable those skilled in the art to fully understand the scope of the present disclosure.

[0019] As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items.

[0020] The terms used herein are only used to describe specific embodiments and are not intended to limit the present disclosure. As used herein, the singular forms "a", "an" and "the" are also intended to include the plural forms, unless the context clearly indicates otherwise. It will also be understood that when the terms "comprising" and / or "made of" are used in this specification, the presence of the features, wholes, steps, operations, elements and / or components is specified, but the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or groups thereof is not excluded.

[0021] The embodiments described herein may be described with reference to plan views and / or cross-sectional views by means of idealized schematic diagrams of the present disclosure. Therefore, the example illustrations may be modified according to manufacturing techniques and / or tolerances. Therefore, the embodiments are not limited to the embodiments shown in the accompanying drawings, but include modifications of the configurations formed based on the manufacturing process. Therefore, the regions illustrated in the accompanying drawings have schematic properties, and the shapes of the regions shown in the figures illustrate the specific shapes of the regions of the elements, but are not intended to be limiting.

[0022] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those of ordinary skill in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant art and the present disclosure, and will not be interpreted as having an idealized or overly formal meaning unless explicitly defined as such herein.

[0023] The present disclosure provides a method for detecting compliance of an installation package. Figure 1 As shown, the method comprises the following steps:

[0024] Step S11, when it is detected that the installation package contains an image during the installation process of the installation package, the image is analyzed to obtain first description information of the image, where the first description information is text information in human language that represents the association between elements in the image.

[0025] The installation package includes but is not limited to APK (Android application package). In the disclosed embodiment, the analysis and prediction of the image content in the installation package is started during the installation phase of the installation package, that is, the first description information of the image in the installation package is obtained using a pre-trained image analysis model. In this step, if the installation command of the installation package is detected, it is detected during the installation process of the installation package whether the installation package contains an image. If the image is contained, the image is input into the pre-trained image analysis model, and the image is analyzed using the image analysis model to obtain text information in human language that characterizes the correlation between the elements in the image, that is, the first description information.

[0026] There are many methods for detecting whether an installation package contains an image. For example, the method may be to detect whether an image is contained by detecting a wildcard in a file directory of the installation package.

[0027] In some embodiments, when it is detected that the installation package contains an image and the image is complete during the installation process of the installation package, the image is analyzed to obtain the first description information of the image. That is, before analyzing the image, it is first detected whether the image is complete. If the image is incomplete, such as damaged, there is no need to perform image analysis. Even if the image analysis is performed, the first description information cannot be obtained due to the missing image. Therefore, detecting the integrity of the image before analyzing the image can avoid performing invalid image analysis operations.

[0028] Step S12: determining compliance of the installation package according to the first description information.

[0029] The first description information is text information, and text comparison can be used to determine whether the first description information contains sensitive information to determine whether the installation package is compliant. It should be noted that the first description information can not only reflect the relevance of the elements in the image, but also reflect the subject's actions and behaviors. The first description information can be in the form of sentences, phrases, etc. Therefore, when detecting the compliance of the installation package, detection based on phrases can improve accuracy compared to compliance detection based on single keywords.

[0030] The installation package compliance detection method provided by the embodiment of the present disclosure detects whether the installation package contains an image during the installation process of the installation package. If the image is contained, the image is analyzed to obtain the first description information of the image, that is, the text information in human language that characterizes the correlation between the elements in the image, and the compliance of the installation package is determined based on the first description information; the embodiment of the present disclosure can automatically analyze whether the image in the installation package complies with the regulations during the installation process of the installation package, and advance the timing of the compliance detection of the application, so that the user can be prompted with an alarm in advance to reduce the security risks caused by non-compliant applications. Moreover, the first description information can include phrases that can reflect information such as behavioral actions. Therefore, compliance detection based on the first description information can improve the accuracy of detection.

[0031] In the embodiment of the present disclosure, an image is analyzed using an image analysis model, such as Figure 2 As shown, the step of analyzing the image to obtain the first description information of the image (i.e., step S11) includes the following steps:

[0032] Step S111, identifying elements in the image and determining the attributes of each element.

[0033] The elements in the image may include entities, and may include people, animals, objects, etc. by way of example. The elements may also include actions of the "human" entity, and may include "holding", "playing", etc. by way of example. In the disclosed embodiments, the attributes of the elements refer to the sentence component attributes of the elements, and the sentence component attributes may include: subject, predicate, object, etc. The attributes of entity elements are usually subjects and objects, and the attributes of action elements are usually predicates. The sentence component attributes also determine the position and order of the elements in the first description information to a certain extent. In this step, the image in the detected installation package is input into the image analysis model, and the image analysis model identifies the elements and their attributes of the image.

[0034] Step S112, determining the association weights between the elements.

[0035] The image analysis model calculates the association weight between any two elements. The association weight represents the degree of association between the elements. The association weight takes a value of (0,1). The larger the value, the greater the association between the two elements.

[0036] Step S113: Generate first description information of the image according to the attributes of each element, the association weights between the elements and the grammatical rules of human language.

[0037] According to the grammatical rules of human language and the attributes of each element, the position and order of the elements in the first description information can be roughly determined, and the position of the elements can be adjusted in combination with the association weights between the elements, thereby obtaining the final first description information. Human language refers to languages ​​of different language families, such as Chinese, English, French, Spanish, etc. The embodiments of the present disclosure can generate first description information in different languages.

[0038] Assume that the elements in the image are identified as "female", "holding", "playing cards", and "playing", the attributes of "female" and "playing cards" are subjects or objects, and the attributes of "holding" and "playing" are "predicates". The association weights between the elements are shown in Table 1:

[0039] Table 1

[0040] female Holding Poker play female 0.9 0.7 0.8 0.4 Holding 0.7 0.9 0.8 0.1 Poker 0.8 0.8 0.9 0.7 play 0.4 0.1 0.7 0.9

[0041] Through Table 1, it can be determined that the correlation weights of the three elements "female", "holding" and "playing cards" are relatively large. Combining the attributes of these three elements and the grammatical rules of Chinese, the first descriptive information of "female holding playing cards" can be obtained.

[0042] The image analysis model of the embodiment of the present disclosure is trained for "text-image pairs", combining one-dimensional text and two-dimensional images to achieve polymorphic deep learning, and present the content in the image through language description. For example, by identifying the elements including "playing cards" and "women" in the image through semantic segmentation, the image analysis model can describe the content in the image through human language, that is, output the first description information such as "women holding playing cards". The first description information can reflect the action of the subject and the correlation between the elements. In the process of training "text-image pairs", the attention mechanism is used. The attention mechanism calculates the correlation between each word as a weight during training, so that the relationship between the constituent elements of the sentence has strong correlation and weak correlation, so that the image analysis model trained in this way can also have the same effect in the application stage. In the related art, only static elements in the image can be detected, and the correlation between the static elements cannot be given. Therefore, the embodiment of the present disclosure uses the first description information to represent the image, which is more intuitive and flexible. The attenion mechanism emphasizes the subject behavior. The “text-image pair” dataset combined with the attenion mechanism also has the advantage of zero shot (zero-shot learning), and labels (elements) that have never appeared in the training set can also be predicted. For example, there are two animals in the training set: horses and tigers (with stripe features). If the current image is a zebra, the feature of the zebra is a combination of the feature of the horse and the feature of the tiger's stripes, but the training set does not have this combined feature. Based on zero shot, the combined features that are not in the training set can be predicted, thereby predicting "zebra". Zero shot can predict more image elements when training data is limited, thereby improving generalization ability.

[0043] In the image analysis model training phase, contrastive learning of "text-image pairs" can be used to train the image analysis model. Specifically, feature extraction is performed on the image and text respectively. The framework of image feature extraction is not limited. For example, the Resnet model can be used to extract image features. The model used for text feature extraction is not limited. For example, the Bert model can be used to extract text features. The Bert model extracts three features from the text, namely, word vector, classification information, and position information. The image feature and the text feature are directly multiplied to calculate the cosine distance, that is, the cosine distance between the same text-image pairs is 1, and the cosine distance between different pairs of text-image is 0. The cosine distance is used to determine whether a "text-image pair" can be formed. A large number of sensitive images and the text information corresponding to the sensitive images form text-image pairs, and further form a training data set.

[0044] In some embodiments, Figure 3 As shown, the step of determining the compliance of the installation package according to the first description information (i.e., step S12) includes the following steps:

[0045] Step S121, detecting whether the first description information includes sensitive information.

[0046] Step S122: When the first description information includes sensitive information, determine that the installation package is non-compliant.

[0047] In the disclosed embodiment, sensitive information may include, but is not limited to, drug-related information, gambling-related information, pornography-related information, financial fraud information, and other information that violates laws and regulations. In this step, the first description information may be split into multiple keywords, and each keyword may be used to compare the sensitive information vocabulary, for example, by calculating the matching degree for comparison. If the matching degree is large, the comparison is successful, indicating that the keyword matches the sensitive information, that is, the first description information includes sensitive information, and accordingly, the installation package is not compliant. In addition, the first description information may be in the form of sentences, phrases, etc. In this step, the phrases may also be matched with phrases in the sensitive information vocabulary based on the phrases. If the match is successful, it indicates that the first description information includes sensitive information, and accordingly, it is determined that the installation package is not compliant. Exemplarily, the first description information may include the phrase "playing poker cards", and the sensitive information vocabulary may be directly compared based on the phrase, while in the related art, the sensitive information vocabulary can only be compared based on the keyword "playing poker cards". Obviously, the sensitive information vocabulary based on the phrase is more accurate.

[0048] Step S123: Generate first alarm information including first description information, and send the first alarm information.

[0049] If the installation package is not compliant, a first warning message including the first description information is sent to the user, for example, the first warning message is displayed on the terminal. For example, the first warning message may be: "The installation package contains an image of 'a woman holding poker cards', which may lead to gambling, fraud and other risks."

[0050] The first description information is text information in human language that represents the correlation between the elements in the image, and the first warning information includes the first description information. In this way, the first warning information informs the user in a human language that can reflect the correlation between the elements in the image, which is easy for the user to understand intuitively, helps the user to perceive the non-compliant scenes represented by the images in the installation package, and is easier for the user to understand and accept.

[0051] In some embodiments, Figure 4As shown, the generating of the first warning information including the first description information and sending the first warning information (ie, step S123) includes the following steps:

[0052] Step S1231, when the installation package has been installed, detect the activation status of the application corresponding to the installation package. If the application is in the activated state, execute step S1232; if the application is in the inactivated state, execute step S1233.

[0053] Step S1232: Generate first alarm information including first description information, and send the first alarm information.

[0054] Step S1233: Generate first warning information including first description information, save the first warning information, and send the first warning information when the application is activated next time.

[0055] It can be seen from the above steps S1231-1233 that if the corresponding application has been installed when the installation package is detected to be non-compliant, an alarm will be issued when the user activates the application.

[0056] In some embodiments, the generating of the first alarm information including the first descriptive information and sending the first alarm information (i.e., step S123) include the following steps: when the installation package has not completed the installation, generating the first alarm information including the first descriptive information and sending the first alarm information. That is to say, if it is detected during the installation process of the installation package that the image in the installation package includes sensitive information and the application corresponding to the installation package has not been installed, a direct alarm is issued. The disclosed embodiment performs content analysis on the image in the installation package of the application based on a deep learning model, represents the content displayed by the image in text form, and informs the user through an alarm message, which will be more intuitive, easier for users to accept, and more clear on the basis of the reasons why the installation package is considered non-compliant, thereby improving the user's trust and acceptance of the detection results.

[0057] In some embodiments, Figure 5 As shown, the installation package compliance detection method may also include the following steps:

[0058] Step S21, when it is detected that the installation package contains an image during the installation process of the installation package, identifying text information in the image.

[0059] If the installation package contains an image and there is text in the image, the text in the image needs to be recognized. Specifically, the area where the text is located is predicted, and the text information is segmented in the area to obtain multiple words. The text information is angle-corrected by calculating the similarity between two adjacent words to recognize abnormal fonts (for example, artistic fonts, italics, etc.).

[0060] It should be noted that, for the text files in the installation package, it is also necessary to identify the text information in the text files. For example, OCR (Optical Character Recognition) technology can be used to identify the text information in the text files.

[0061] Step S22: if the text information includes sensitive information, generate second warning information and send the second warning information.

[0062] In some embodiments, Figure 6 As shown, after sending the first warning information and / or the second warning information, the installation package compliance detection method may further include the following steps:

[0063] Step S31, marking the application corresponding to the installation package as a target application.

[0064] If the first warning information and / or the second warning information has been sent, it means that the corresponding application has been detected as non-compliant, and the application is marked as a target application, and the target application is a high-risk application.

[0065] Step S32, monitoring the target application during its operation.

[0066] During the operation of the target application, the target application can be further monitored to determine whether it generates sensitive information during the operation.

[0067] It can be seen from steps S31-S32 that the disclosed embodiment targets non-compliant applications found during the installation process, i.e., sensitive information is detected, an alarm message is sent, and the installation of the application is not temporarily suspended, nor is the application blocked from running. In addition, the application is also marked, and the application continues to be monitored during subsequent operation. In other words, when the installation package is detected to be non-compliant before the installation package is installed, only an alarm message is sent, and the installation program will not be forcibly suspended, which can prevent the terminal from causing poor physical examination of the user.

[0068] In some embodiments, the monitoring of the target application during the operation of the target application (i.e., step S32) includes the following steps: when the target application generates a target image during operation, the target image is analyzed to obtain second description information of the target image, the second description information being text information in human language that characterizes the association between elements in the image; and the compliance of the installation package is determined according to the second description information. Wherein, the determination of the compliance of the installation package according to the second description information includes the following steps: detecting whether the second description information includes sensitive information; when the second description information includes sensitive information, determining that the installation package is non-compliant; generating third warning information including the second description information, and sending the third warning information. That is, for the target application whose installation package has been detected to include sensitive information during the installation phase, the generated image (i.e., the target image) during the operation of the target application is detected to determine whether there is sensitive information in the target image. If there is, a warning message that is easy for the user to understand is generated to warn, thereby achieving further monitoring of the target application during the application operation phase.

[0069] In some embodiments, after detecting whether sensitive information is included in the second description information, the installation package compliance detection method may further include the following steps: when sensitive information is included in the second description information, recording the number of sensitive information detected in the second description information; when the number is greater than a preset threshold, executing a preset target operation. It should be noted that the number refers to the total number of identical or different sensitive information. For example, if the same sensitive information appears multiple times, the number of sensitive information also needs to be counted multiple times. In other words, during the operation of the target application, if a large amount of sensitive information appears and / or the same sensitive information appears repeatedly, further measures are automatically triggered. Exemplarily, the further measures include but are not limited to alarming, disabling the target application, deleting the target application, etc.

[0070] The disclosed embodiment starts analyzing and detecting the image resource content in the installation package during the installation phase of the installation package, and when sensitive information is detected, alerts the user by using text information in human language (i.e., first description information) that represents the correlation between elements in the image. The first description information can emphasize the correlation between the action and the subject, so the warning information including the first description information is easier for the user to understand and accept. The image analysis model used for image analysis is trained based on "text-image pairs", a kind of deep learning based on polymorphism, combining one-dimensional text and two-dimensional images to achieve polymorphic deep learning, and present the content in the image through language description.

[0071] The disclosed embodiment can automatically obtain image resources in the installation package and automatically analyze whether the content in the image is compliant, thereby saving human resources, eliminating the need for a large amount of manpower for supervision and inefficient troubleshooting, and improving the efficiency of application compliance detection.

[0072] like Figure 7 As shown, the embodiment of the present disclosure also provides an installation package compliance detection device, which includes: one or more processors 701, a storage device 702 and one or more I / O interfaces 703; wherein, one or more programs are stored on the storage device 702, and when the one or more programs are executed by the one or more processors, the one or more processors 701 implement the installation package compliance detection method provided in the aforementioned embodiments; the I / O interface 703 is connected between the processor 701 and the storage device 702, and is configured to implement information interaction between the processor 701 and the storage device 702.

[0073] Among them, the processor 701 is a device with data processing capabilities, including but not limited to a central processing unit (CPU); the storage device 702 is a device with data storage capabilities, including but not limited to random access memory (RAM, more specifically SDRAM, DDR, etc.), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory (FLASH); the I / O interface (read-write interface) 703 is connected between the processor 701 and the storage device 702, and can realize information interaction between the processor 701 and the storage device 702, including but not limited to a data bus (Bus), etc.

[0074] The embodiments of the present disclosure further provide a computer-readable medium on which a computer program is stored, wherein when the computer program is executed, the installation package compliance detection method provided in the aforementioned embodiments is implemented.

[0075] It will be appreciated by those skilled in the art that all or some of the steps in the method disclosed above, the functional modules / units in the device may be implemented as software, firmware, hardware, and appropriate combinations thereof. In hardware implementations, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed by several physical components in cooperation. Some physical components or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or temporary medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, it is well known to those of ordinary skill in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0076] Example embodiments have been disclosed herein, and although specific terms are employed, they are used and should be interpreted only in a general illustrative sense and not for limiting purposes. In some instances, it will be apparent to those skilled in the art that, unless otherwise expressly stated, features, characteristics, and / or elements described in conjunction with a particular embodiment may be used alone or in combination with features, characteristics, and / or elements described in conjunction with other embodiments. Therefore, it will be understood by those skilled in the art that various changes in form and detail may be made without departing from the scope of the invention as set forth in the appended claims.

Claims

1. A method for detecting compliance of an installation package, characterized in that: include: When it is detected during the installation process of the installation package that the installation package contains an image, the image is analyzed to obtain first description information of the image, where the first description information is text information in human language that represents the association between elements in the image; The compliance of the installation package is determined according to the first description information.

2. The method according to claim 1, characterized in that The step of analyzing the image to obtain first description information of the image includes: identifying elements in the image and determining properties of each of the elements; Determining the association weights between the elements; The first description information of the image is generated according to the attributes of each of the elements, the association weights between the elements and the grammatical rules of the human language.

3. The method according to claim 1, characterized in that The determining the compliance of the installation package according to the first description information includes: Detecting whether the first description information includes sensitive information; In a case where the first description information includes sensitive information, determining that the installation package is non-compliant; Generate first alarm information including the first description information, and send the first alarm information.

4. The method according to claim 3, characterized in that The generating the first warning information including the first description information and sending the first warning information includes: When the installation package has been installed, detecting the activation status of the application corresponding to the installation package; When the application is in an enabled state, generating first warning information including the first description information, and sending the first warning information; When the application is in an inactive state, first warning information including the first description information is generated, the first warning information is saved, and the first warning information is sent when the application is activated.

5. The method according to claim 3, characterized in that The generating the first warning information including the first description information and sending the first warning information includes: In the case that the installation package has not been installed completely, first warning information including the first description information is generated, and the first warning information is sent.

6. The method according to claim 3, characterized in that The method further comprises: When it is detected during the installation process of the installation package that the installation package contains an image, identifying text information in the image; In the case where the text information includes sensitive information, second warning information is generated and sent.

7. The method according to claim 6, characterized in that After sending the first warning information and / or the second warning information, the method further includes: Marking the application corresponding to the installation package as a target application; The target application is monitored during the execution of the target application.

8. The method according to claim 7, characterized in that The monitoring of the target application during the running of the target application includes: When the target application generates a target image during operation, analyzing the target image to obtain second description information of the target image, where the second description information is text information in human language that represents the correlation between elements in the image; The compliance of the installation package is determined according to the second description information.

9. The method according to claim 8, characterized in that The determining the compliance of the installation package according to the second description information includes: Detecting whether the second description information includes sensitive information; In a case where the second description information includes sensitive information, determining that the installation package is non-compliant; Generate third warning information including the second description information, and send the third warning information.

10. The method according to claim 9, characterized in that After detecting whether the second description information includes sensitive information, the method further includes: In the case where the second description information includes sensitive information, recording the amount of sensitive information detected in the second description information; When the number is greater than a preset threshold, a preset target operation is performed.

11. The method according to any one of claims 1 to 10, characterized in that: When it is detected during the installation process of the installation package that the installation package contains an image and the image is complete, the image is analyzed to obtain first description information of the image.

12. An installation package compliance detection device, comprising: one or more processors; a storage device having one or more programs stored thereon; When the one or more programs are executed by the one or more processors, the one or more processors implement the installation package compliance detection method according to any one of claims 1 to 11; One or more I / O interfaces are connected between the processor and the storage device and are configured to implement information interaction between the processor and the storage device.

13. A computer readable medium having a computer program stored thereon, wherein: When the program is executed, the installation package compliance detection method according to any one of claims 1 to 11 is implemented.

Images