Malicious software detection method, device, equipment and product

By analyzing the frequency of API operation and folder path frequency between the malware, the characteristics are fused to improve the accuracy of malware detection, the problem of difficulty in feature extraction in the existing technology is solved, and more accurate malware classification and behavior are achieved. Characteristics.

CN118965346AActive Publication Date: 2024-11-15XIDIAN UNIV HANGZHOU RES INST

Patent Information

Application Number
CN202411000175.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-24
Publication Date
2024-11-15
Estimated Expiration
2044-07-24

AI Technical Summary

Technical Problem

The existing malware classification algorithms are difficult to extract effective feature quantities and cannot fully reflect the behavioral characteristics of the malware family, resulting in inaccurate classification.

Method used

By obtaining the family tags of the malware to be detected, perform similarity analysis of operation frequency between application programming interface groups and folder path operation frequency analysis, integrate concentrated distance characteristics and operation frequency characteristics, and use the trained malware detection model to determine it category.

Benefits of technology

Improve the accuracy of malware detection, which can better reflect the behavioral characteristics of the malware family, help understand the attacker's behavior and motivation, and enhance the effectiveness of security measures.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118965346A_ABST
    Figure CN118965346A_ABST
Patent Text Reader

Abstract

The invention discloses a malicious software detection method, device, equipment and product, and relates to the technical field of network security, the method comprises the following steps: obtaining a family label of to-be-detected malicious software; performing similarity analysis processing on the operation frequency between the application programming interface groups on the to-be-detected malicious software to obtain the similarity between the application programming interface operations, and determining the similarity characteristics of the to-be-detected malicious software based on the similarity; performing folder path operation analysis processing on the to-be-detected malicious software to obtain folder operation frequency characteristics of the to-be-detected malicious software; performing feature fusion on the concentrated distance feature and the operation frequency feature to obtain a fusion feature; and determining the category of the to-be-detected malicious software based on the fusion feature. According to the method, the type of the malicious software can be more effectively detected, and the detection accuracy is improved, so that the characteristics of the malicious software can be better known, and maintenance personnel can take more effective security measures conveniently.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of network security technology, and in particular to a method, device, equipment and product for detecting malicious software. Background Art

[0002] In the field of malware classification based on machine learning, a variety of high-precision and high-efficiency classification algorithms have been proposed. However, the existing classification algorithms have difficulty extracting the feature quantities of malware and cannot fully reflect the behavioral characteristics of malware families. It can be seen that how to accurately classify malware is an important issue that needs to be urgently solved in the industry. Summary of the invention

[0003] In view of this, the embodiments of the present invention provide a malware detection method, apparatus, device and product to solve the problem in the prior art that it is difficult to extract feature quantities of malware and that the behavioral characteristics of malware families cannot be fully reflected.

[0004] According to a first aspect, an embodiment of the present invention provides a method for detecting malware, the method comprising: Get the family label of the malware to be detected; Performing similarity analysis on the operation frequencies of application programming interface groups of the malware to be detected, obtaining similarities between various application programming interface operations, and determining similarity features of the malware to be detected based on the similarities; Performing folder path operation analysis on the malware to be detected to obtain folder operation frequency characteristics of the malware to be detected; Perform feature fusion on the concentrated distance feature and the operation frequency feature to obtain a fusion feature; Determine the category of the malware to be detected based on the fused features.

[0005] In combination with the first aspect, in a first implementation of the first aspect, obtaining the family label of the malware to be detected specifically includes: Get analysis reports of malware to be detected; Extract the family label of the detected malware from the analysis report.

[0006] In combination with the first aspect, in the second implementation of the first aspect, the similarity analysis and processing of the operation frequencies between application programming interface groups of the malware to be detected is performed to obtain the similarity between the various application programming application programming interface operations, and the similarity characteristics of the malware to be detected are determined based on the similarity, specifically including: Perform application programming interface operation frequency analysis on the malware to be detected to obtain the calling frequency of each application programming operation; The similarities between the programming operations of each application are obtained to obtain a similarity matrix between the programming operations of each application, and the similarity matrix is ​​converted into a condensed distance feature, and the similarity feature is the condensed distance matrix obtained by the conversion.

[0007] In combination with the first aspect, in a third implementation of the first aspect, the performing of folder path operation analysis and processing on the malware to be detected to obtain folder operation frequency characteristics of the malware to be detected specifically includes: Get the folder path where the malware to be detected has been operated; The operation frequencies of the folder paths operated by the malware to be detected are counted, and folder operation frequency features are generated based on the operation frequencies.

[0008] In combination with the first aspect, in a fourth implementation of the first aspect, in determining the category of the malware to be detected based on the fusion features, the category is obtained through a trained malware detection model.

[0009] In combination with the fourth implementation of the first aspect, in the fifth implementation of the first aspect, the input data of the malware detection model includes the fusion features and family labels of the malware to be detected; the malware detection model is based on samples, the fusion features corresponding to the samples, and the family labels, and is obtained after training based on supervised machine learning methods.

[0010] In combination with the fifth implementation of the first aspect, in a sixth implementation of the first aspect, the training process of the malware detection model includes: Obtaining the operation frequency characteristics and operation frequency characteristics of the malware to be detected, and performing feature fusion on the concentrated distance characteristics and the operation frequency characteristics to obtain fusion characteristics; Extract the family label of the malware to be detected; The fused features are used as input data for training, and the family labels are used as training labels for the fused features. The supervised machine learning method is used for training to obtain a malware detection model for outputting the category of the malware to be detected.

[0011] According to a second aspect, an embodiment of the present invention further provides a malware detection device, the device comprising: The label extraction module is used to obtain the family label of the malware to be detected; The first analysis module is used to perform similarity analysis on the operation frequencies of application programming interface groups of the malware to be detected, obtain the similarity between the operations of various application programming application programming interfaces, and determine the similarity characteristics of the malware to be detected based on the similarity; The second analysis module is used to perform folder path operation analysis on the malware to be detected to obtain folder operation frequency characteristics of the malware to be detected; A feature fusion module is used to fuse the concentrated distance feature and the operation frequency feature to obtain a fused feature; The category detection module is used to determine the category of the malware to be detected based on the fusion features.

[0012] According to a third aspect, an embodiment of the present invention further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of any of the malware detection methods described above are implemented.

[0013] According to a third aspect, an embodiment of the present invention further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of any of the malware detection methods described above are implemented.

[0014] According to a fourth aspect, an embodiment of the present invention further provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of the malware detection method as described in any one of the above items.

[0015] The malware detection method, device, equipment and product of the present invention perform similarity analysis on the operation frequencies of API groups of the malware to be detected to obtain the similarity features of the malware to be detected, then perform folder path operation analysis on the malware to be detected to obtain the folder operation frequency features of the malware to be detected, and then perform feature fusion on the concentrated distance features and the operation frequency features to obtain fused features. Since the similarity between API groups can reflect the behavior pattern of the malware family, that is, the concentrated distance feature can better reflect the behavior pattern of the malicious detection to be detected, and the folder path operation frequency can reflect the representative behavior of the malware family, that is, the folder operation frequency feature can better reflect the representative behavior of the malicious detection to be detected, the fused feature can more accurately reflect the behavior characteristics of the malware family to be detected, and then the category of the malware to be detected determined by using the fused feature is more accurate, and the category of the malware can be detected more effectively, thereby improving the accuracy of detection, thereby better understanding the characteristics of the malware, and then better understanding the behavior and motivation of the attacker, and facilitating maintenance personnel to take more effective security measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The features and advantages of the present invention will be more clearly understood by referring to the accompanying drawings, which are schematic and should not be construed as limiting the present invention in any way. In the accompanying drawings: Figure 1 A schematic diagram showing the flow of the malware detection method provided by the present invention; Figure 2 A specific flow chart of step S10 in the malware detection method provided by the present invention is shown; Figure 3 A specific flow chart of step S20 in the malware detection method provided by the present invention is shown; Figure 4 A specific flow chart of step S30 in the malware detection method provided by the present invention is shown; Figure 5 A schematic diagram showing a malware detection model training process in the malware detection method provided by the present invention is shown; Figure 6 A schematic diagram of the structure of a malware detection device provided by the present invention is shown; Figure 7 It is a schematic diagram of the hardware structure of the electronic device provided in the embodiment of the present application. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0018] In the field of malware classification based on machine learning, a variety of high-precision and high-efficiency classification algorithms have been proposed. However, the existing classification algorithms have difficulty in extracting the features of malware and cannot fully reflect the behavioral characteristics of malware families. For example: In the classification of malware families based on opcode, it is difficult to extract opcodes that can truly reflect behavioral characteristics due to the influence of packing and code obfuscation. In the classification of malware families based on network traffic, the interception of communication traffic and the fact that some malware will not be executed when no network communication is detected will increase the difficulty of extracting the network communication content of malware.

[0019] Therefore, how to accurately classify malware is an important issue that needs to be urgently addressed in the industry.

[0020] In order to solve the above problems, a malware detection method is provided in this specification, aiming to more effectively detect the category of malware. The malware detection method provided in this specification can be applied to electronic devices with suspicious file analysis functions. The electronic devices may include notebooks, desktop computers, smart phones, smart wearable devices (virtual reality glasses, smart watches, etc.), tablet computers, etc. Of course, the malware detection method provided in this specification can also be applied to applications running in the above-mentioned electronic devices. For example, the malware detection method can be applied to a browser with a suspicious file analysis function, and can also be applied to suspicious file processing software. Figure 1 FIG. 1 is a flow chart of a method for detecting malware according to an embodiment of the present invention. Figure 1 As shown, the method may include the following steps: S10: Obtain the family label of the malware to be detected. The family label is used to subsequently determine the category of the malware to be detected.

[0021] S20, performing similarity analysis on the operation frequencies of application programming interface groups of the malware to be detected, obtaining similarities between various application programming interface (Application Programming Interface, API) operations, and determining similarity features of the malware to be detected based on the similarities.

[0022] S30: Perform folder path operation analysis on the malware to be detected to obtain folder operation frequency characteristics of the malware to be detected.

[0023] S40, performing feature fusion on the concentrated distance feature and the operation frequency feature to obtain a fused feature.

[0024] Since the similarity between API groups can reflect the behavior pattern of the malware family, that is, the concentrated distance feature can better reflect the behavior pattern of the malicious detection to be detected, and the folder path operation frequency can reflect the representative behavior of the malware family, that is, the folder operation frequency feature can better reflect the representative behavior of the malicious detection to be detected, the fused feature obtained by feature fusion based on the concentrated distance feature and the operation frequency feature can more accurately reflect the behavior characteristics of the malware family to be detected.

[0025] S50: Determine the category of the malware to be detected based on the fusion features obtained by the above fusion.

[0026] In this embodiment, the category of the malware to be detected can be determined by a trained malware detection model. The malware detection model is used to classify the malware to be detected and output the category corresponding to the malware.

[0027] Malware detection models using machine learning classifiers , such as support vector machine classifier, random forest classifier, classifier based on Bayesian theory, etc. The malware detection model is based on samples, fusion features corresponding to the samples and family labels, and is obtained after training based on supervised machine learning.

[0028] The malware detection method of the present invention performs similarity analysis on the operation frequencies of API groups of the malware to be detected to obtain the similarity features of the malware to be detected, performs folder path operation analysis on the malware to be detected to obtain the folder operation frequency features of the malware to be detected, and then performs feature fusion on the concentrated distance features and the operation frequency features to obtain fused features. Since the similarity between API groups can reflect the behavior pattern of the malware family, that is, the concentrated distance feature can better reflect the behavior pattern of the malicious detection to be detected, and the folder path operation frequency can reflect the representative behavior of the malware family, that is, the folder operation frequency feature can better reflect the representative behavior of the malicious detection to be detected, the fused feature can more accurately reflect the behavior characteristics of the malware family to be detected, and then the category of the malware to be detected determined by using the fused feature is more accurate, and the category of the malware can be more effectively detected, thereby improving the accuracy of detection, thereby better understanding the characteristics of the malware, and then better understanding the behavior and motivation of the attacker, and facilitating maintenance personnel to take more effective security measures.

[0029] Combine the following Figure 2 Describing the malware detection method of the present invention, step S10 specifically includes: S11. Obtain an analysis report of the malware to be detected, for example, obtain the analysis report of the malware to be detected from VirusTotal. The analysis report may include analysis results of up to 70 antivirus engines.

[0030] S12. Use label extraction tools such as Avclass to extract the family label (Ground Truth) of the malware to be detected from the analysis report, and then obtain the family label vector of the malware to be detected. .

[0031] Combine the following Figure 3 Describing the malware detection method of the present invention, step S20 specifically includes: S21. Perform application programming interface operation frequency analysis on the malware to be detected to obtain the calling frequency of each application programming operation.

[0032] Use sandbox tools such as Cuckoo Sandbox to dynamically analyze the malware to be detected and collect the malware in the API group The calling frequency of each API operation in each subgroup, where each subgroup is equivalent to an API operation.

[0033] Specifically, the API group There are five subgroups, namely file operation API group , Cryptographic Operation API Group , Registry Operation API Group , Communication operation API group as well as ,Right now:

[0034] S22, obtaining the similarity between each application programming operation, obtaining a similarity matrix between each application programming operation, and converting the similarity matrix into a condensed distance feature, where the similarity feature is the condensed distance matrix obtained by the conversion.

[0035] After that, use the similarity calculation function Calculate the similarity between each subgroup in the API group G to obtain the API group similarity matrix of the malware to be detected ,in Both represent API groups The API subgroup in .

[0036] Then use the condensed matrix calculation function number Transform the above similarity matrix is the condensed distance matrix. In this way, the condensed distance matrix of the malware to be detected can be obtained, that is, , Indicates malware to be detected.

[0037] Combine the following Figure 4 Describing the malware detection method of the present invention, step S30 specifically includes: S31. Obtain the folder path where the malware to be detected has been operated.

[0038] Use sandbox tools such as Cuckoo Sandbox to dynamically analyze the malware to be detected and collect the folder path collection that it has opened, copied, written, read, and deleted. .

[0039] In this embodiment, the folder where the file operated by the malware is located is also the object of the folder operation path frequency statistics, that is, the file name is removed from the file path character string, and only the path of the folder where the file is located is retained.

[0040] S32. Count the operation frequencies of the folder paths operated by the malware to be detected, and generate folder operation frequency features based on the operation frequencies. The folder operation frequency features are the folder operation frequency matrix.

[0041] Count the above folder path collection The operation frequency of each operation type in each folder path in the folder is obtained to obtain the folder operation frequency vector, that is:

[0042] in, Indicates the opening operation of the malware to be detected; Indicates the copy operation of the malware to be detected; Indicates the write operation of the malware to be detected; Indicates the read operation of the malware to be detected; Indicates the deletion operation of the malware to be detected. Of course, the operation types of the folder path are not limited to the above five types, and the operation types can be deleted according to the specific scenario requirements.

[0043] Then obtain the folder operation frequency matrix for the malware to be detected, namely:

[0044] Combine the following Figure 5 The malware detection method of the present invention is described. The malware detection model is trained in the following manner: Get Sample Analysis reports, such as the above samples obtained from VirusTotal The analysis report of the sample The analysis results.

[0045] Then use label extraction tools such as Avclass to extract samples from the analysis report The family label (GroundTruth) of all samples Family label collection and the family label vector .

[0046] It is understandable that the sample set All samples in They are all sample malware. Various malware can be collected as sample data to provide massive data support for subsequent training.

[0047] The sample malware may be stored in the electronic device in advance, or may be obtained by the electronic device from the outside. There is no restriction on the specific method of obtaining the sample malware. It only needs to ensure that the electronic device can obtain the sample malware. That's it.

[0048] For a sample , use sandbox tools such as Cuckoo Sandbox to Perform dynamic analysis and collect API groups The frequency of each API call in each subgroup.

[0049] After that, use the similarity calculation function Calculate the similarity between each subgroup in the API group G and get the sample Similarity matrix between API groups ,in Both represent API groups The API subgroup in .

[0050] Then use the condensed matrix calculation function Transform the above similarity matrix is the condensed distance matrix, that is:

[0051] In this way, we can obtain the sample set All samples in The condensed distance matrix is , Representation sample The total number of .

[0052] For a sample , and then use sandbox tools such as Cuckoo Sandbox to analyze the sample. Perform dynamic analysis to collect the folder paths that have been opened, copied, written, read, and deleted Sample All the folder paths that have been operated in the sample are merged into .

[0053] Similarly, the folder where the file operated by the malware is located is also the object of the folder operation path frequency statistics, that is, the file name is removed from the file path string, and only the path of the folder where the file is located is retained.

[0054] And for each sample , count the above folder path collection The operation frequency of each operation type in each folder path in the folder is obtained to obtain the folder operation frequency vector, that is:

[0055] in, Representation sample Open operation; Representation sample The copy operation; Representation sample Write operation; Representation sample Read operation; Representation sample Of course, the operation types of folder paths are not limited to the above five types. The operation types can be deleted according to the specific scene requirements.

[0056] Therefore, we can also obtain All samples in The folder operation frequency matrix is:

[0057] In this embodiment, feature fusion is to combine the sample set All samples in The condensed distance matrix With sample collection All samples in Folder operation frequency matrix Perform feature fusion to obtain fusion features. Correspondingly, the fusion features are:

[0058] Number of features

[0059] Malware detection models using machine learning classifiers , such as support vector machine classifier, random forest classifier, classifier based on Bayesian theory, etc. The data used for training includes: fusion features and family label vector Then supervised machine learning was used to validate the machine learning classifier based on 10-fold cross validation. Perform parameter optimization until the optimal parameter set is found .

[0060] Train the malware detection model using the optimal set of parameters Classify the malware to be detected into families and obtain the predicted labels of the malware to be detected ,Right now:

[0061] Specifically, the malware model first obtains the fusion features of the malware to be detected, then obtains the family labels of the malware to be detected, and then uses the fusion features as input data for training, and uses the family labels as training labels for the fusion features, and uses supervised machine learning methods for training.

[0062] The following is a description of a malware detection device provided by an embodiment of the present invention. The malware detection device described below and the malware detection method described above can be referenced to each other.

[0063] In order to solve the above problems, a malware detection device is provided in this specification, aiming to obtain a more realistic and accurate real resistance value of the resistor. Figure 6 is a schematic diagram of the structure of a malware detection device according to an embodiment of the present invention. Figure 6 As shown, the device may include: The label extraction module 10 is used to obtain the family label of the malware to be detected. The family label is used to subsequently determine the category of the malware to be detected.

[0064] The first analysis module 20 is used to perform similarity analysis on the operation frequencies between application programming interface groups of the malware to be detected, obtain the similarity between various API operations, and determine the similarity characteristics of the malware to be detected based on the similarity.

[0065] The second analysis module 30 is used to perform folder path operation analysis on the malware to be detected to obtain folder operation frequency characteristics of the malware to be detected.

[0066] The feature fusion module 40 is used to fuse the concentrated distance feature and the operation frequency feature to obtain a fused feature.

[0067] Since the similarity between API groups can reflect the behavior pattern of the malware family, that is, the concentrated distance feature can better reflect the behavior pattern of the malicious detection to be detected, and the folder path operation frequency can reflect the representative behavior of the malware family, that is, the folder operation frequency feature can better reflect the representative behavior of the malicious detection to be detected, the fused feature obtained by feature fusion based on the concentrated distance feature and the operation frequency feature can more accurately reflect the behavior characteristics of the malware family to be detected.

[0068] The category detection module 50 is used to determine the category of the malware to be detected based on the fusion features obtained by the above fusion.

[0069] In this embodiment, the category of the malware to be detected can be determined by a trained malware detection model. The malware detection model is used to classify the malware to be detected and output the category corresponding to the malware.

[0070] Malware detection models using machine learning classifiers , such as support vector machine classifier, random forest classifier, classifier based on Bayesian theory, etc. The malware detection model is based on samples, fusion features corresponding to the samples and family labels, and is obtained after training based on supervised machine learning.

[0071] The malware detection device of the present invention performs similarity analysis on the operation frequencies between API groups of the malware to be detected to obtain the similarity features of the malware to be detected, performs folder path operation analysis on the malware to be detected to obtain the folder operation frequency features of the malware to be detected, and then performs feature fusion on the concentrated distance features and the operation frequency features to obtain fused features. Since the similarity between API groups can reflect the behavior pattern of the malware family, that is, the concentrated distance feature can better reflect the behavior pattern of the malicious detection to be detected, and the folder path operation frequency can reflect the representative behavior of the malware family, that is, the folder operation frequency feature can better reflect the representative behavior of the malicious detection to be detected, the fused feature can more accurately reflect the behavior characteristics of the malware family to be detected, and then the category of the malware to be detected determined by using the fused feature is more accurate, and the category of the malware can be detected more effectively, thereby improving the accuracy of detection, thereby better understanding the characteristics of the malware, and then better understanding the behavior and motivation of the attacker, and facilitating maintenance personnel to take more effective security measures.

[0072] Figure 7 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 7 As shown, the electronic device may include: a processor 710 (processor), a communication interface 720 (Communications Interface), a memory 730 (memory) and a communication bus 740, wherein the processor 710, the communication interface 720, and the memory 730 communicate with each other through the communication bus 740. The processor 710 may call the logic command in the memory 730 to execute the malware detection method, which includes: Get the family label of the malware to be detected; Performing similarity analysis on the operation frequencies of application programming interface groups of the malware to be detected, obtaining similarities between various application programming interface operations, and determining similarity features of the malware to be detected based on the similarities; Performing folder path operation analysis on the malware to be detected to obtain folder operation frequency characteristics of the malware to be detected; Perform feature fusion on the concentrated distance feature and the operation frequency feature to obtain a fusion feature; Determine the category of the malware to be detected based on the fused features.

[0073] In addition, the logic instructions in the above-mentioned memory 730 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present invention can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memor), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0074] On the other hand, the present invention further provides a computer program product, the computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions, when the program instructions are executed by a computer, the computer can execute the malware detection method provided by the above methods, the method comprising: Get the family label of the malware to be detected; Performing similarity analysis on the operation frequencies of application programming interface groups of the malware to be detected, obtaining similarities between various application programming interface operations, and determining similarity features of the malware to be detected based on the similarities; Performing folder path operation analysis on the malware to be detected to obtain folder operation frequency characteristics of the malware to be detected; Perform feature fusion on the concentrated distance feature and the operation frequency feature to obtain a fusion feature; Determine the category of the malware to be detected based on the fused features.

[0075] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method for detecting malware provided above is implemented, and the method comprises: Get the family label of the malware to be detected; Performing similarity analysis on the operation frequencies of application programming interface groups of the malware to be detected, obtaining similarities between various application programming interface operations, and determining similarity features of the malware to be detected based on the similarities; Performing folder path operation analysis on the malware to be detected to obtain folder operation frequency characteristics of the malware to be detected; Perform feature fusion on the concentrated distance feature and the operation frequency feature to obtain a fusion feature; Determine the category of the malware to be detected based on the fused features.

[0076] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0077] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0078] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for detecting malware, characterized in that: The method comprises: Get the family label of the malware to be detected; Performing similarity analysis on the operation frequencies of application programming interface groups of the malware to be detected, obtaining the similarity between the operations of various application programming application programming interfaces, and determining similarity features of the malware to be detected based on the similarity; Performing folder path operation analysis on the malware to be detected to obtain folder operation frequency characteristics of the malware to be detected; Perform feature fusion on the concentrated distance feature and the operation frequency feature to obtain a fusion feature; Determine the category of the malware to be detected based on the fused features.

2. The method for detecting malware according to claim 1, characterized in that: The step of obtaining the family label of the malware to be detected specifically includes: Get analysis reports of malware to be detected; Extract the family label of the detected malware from the analysis report.

3. The method for detecting malware according to claim 1, characterized in that: The similarity analysis of the operation frequencies between application programming interface groups of the malware to be detected is performed to obtain the similarity between the operations of each application programming interface, and the similarity characteristics of the malware to be detected are determined based on the similarity, specifically including: Perform application programming interface operation frequency analysis on the malware to be detected to obtain the calling frequency of each application programming operation; The similarities between the programming operations of each application are obtained to obtain a similarity matrix between the programming operations of each application, and the similarity matrix is ​​converted into a condensed distance feature, and the similarity feature is the condensed distance matrix obtained by the conversion.

4. The method for detecting malware according to claim 1, characterized in that: The folder path operation analysis and processing of the malware to be detected is performed to obtain the folder operation frequency characteristics of the malware to be detected, specifically including: Get the folder path where the malware to be detected has been operated; The operation frequency of the folder path operated by the malware to be detected is counted, and the folder operation frequency feature is generated based on the operation frequency.

5. The method for detecting malware according to claim 1, characterized in that: In determining the category of the malware to be detected based on the fusion features, the category is obtained through a trained malware detection model.

6. The method for detecting malware according to claim 5, characterized in that: The input data of the malware detection model includes the fusion features and family labels of the malware to be detected; the malware detection model is obtained after training based on samples, the fusion features and family labels corresponding to the samples, and based on supervised machine learning.

7. The method for detecting malware according to claim 6, characterized in that: The training process of the malware detection model includes: Obtaining the operation frequency characteristics and operation frequency characteristics of the malware to be detected, and performing feature fusion on the concentrated distance characteristics and the operation frequency characteristics to obtain fusion characteristics; Extract the family label of the malware to be detected; The fused features are used as input data for training, and the family labels are used as training labels for the fused features. The supervised machine learning method is used for training to obtain a malware detection model for outputting the category of the malware to be detected.

8. A malware detection device, characterized in that: The device comprises: The label extraction module is used to obtain the family label of the malware to be detected; The first analysis module is used to perform similarity analysis on the operation frequencies of application programming interface groups of the malware to be detected, obtain the similarity between the operations of various application programming application programming interfaces, and determine the similarity characteristics of the malware to be detected based on the similarity; The second analysis module is used to perform folder path operation analysis on the malware to be detected to obtain folder operation frequency characteristics of the malware to be detected; A feature fusion module is used to fuse the concentrated distance feature and the operation frequency feature to obtain a fused feature; The category detection module is used to determine the category of the malware to be detected based on the fusion features.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the malware detection method according to any one of claims 1 to 7 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the malware detection method according to any one of claims 1 to 7 are implemented.

Citation Information

Patent Citations

  • Malicious software family detection method and device

    CN111460448A

  • Feature fusion malicious code family classification method and device and storage medium

    CN113935033A

  • Malicious software family classification method based on multi-dimensional feature fusion and model integration

    CN114386511A

  • Malicious software detection method and device, electronic equipment, medium and product

    CN114491523A

  • Malware family label correction method and device based on hybrid analysis

    CN114936366A

Cited By

  • Method for application type detection and non-transitory machine-readable medium

    TWI938952B