CN102521496A - Method and system for acquiring importance levels of evaluation indexes - Google Patents

Abstract

The present invention provides a method and system for acquiring importance levels of evaluation indicators. The method includes: storing the scoring results obtained by m systems on n evaluation indicators X ₁ , X ₂ ,...,X _i ,...,X _n in the data set S, and saving each system The risk level of the risk level is stored in the data set R, and the entropy H(R) of the risk level in the data set R is calculated; after the information gain rate of n evaluation indicators is obtained, the obtained information gain rate is sorted to obtain n evaluation indicators Importance levels of indicators X ₁ , X ₂ , . . . , Xi _, . . . , X _n .

Description

Method and system for obtaining importance levels of evaluation indicators

technical field

The present invention relates to the field of information processing, in particular to a method and system for acquiring importance levels of evaluation indicators.

Background technique

Information security risk assessment is to use scientific methods and means from the perspective of risk management to systematically analyze the threats and vulnerabilities faced by information systems, assess the degree of harm that may be caused by security incidents, and propose targeted defense measures. Threat protection countermeasures and corrective actions. In the evaluation process, according to the requirements of different industries, the evaluation index and evaluation content of the information system of the industry are put forward, and the final risk level of the system is calculated according to the evaluation index value.

Different industries have different evaluation indicators and evaluation content. Based on the evaluation of different systems in multiple industries, the industry data is comprehensively analyzed, and the laws and commonalities are found from the industry data to provide a new level of importance of industry risk points. Judgment standards and methods, so as to guide the rectification work of information system security construction. Based on the data obtained from the evaluation of different systems in the same industry in the unified evaluation index system and the calculation of the risk level, it is found from the data which evaluation indicators provide a relatively large amount of information for the risk level, that is, which attributes are more important, and there is currently no relevant technology.

Contents of the invention

The present invention provides a method and system for acquiring the importance level of evaluation indicators, and the technical problem to be solved is how to accurately determine the importance level of evaluation indicators for customers.

In order to solve the problems of the technologies described above, the present invention provides the following technical solutions:

A method for obtaining the importance level of evaluation indicators, including:

Save the scoring results of m systems on n evaluation indicators X ₁ , X ₂ ,...,X _i ,...,X _n in the data set S, and save the risk level of each system in the data set R;

Calculate the entropy H(R) of the risk level in the data set R;

Perform steps A~C for each evaluation indicator, including:

Step A. According to the scoring results of the evaluation index _Xi in the data set S, all the risk levels corresponding to the system with the same scoring results for the evaluation index _Xi in the data set R are used as the evaluation index _Xi in the data set R For a corresponding sub-set, by calculating the entropy of the risk level in each sub-set corresponding to the evaluation index X _i , the weighted value of the risk level entropy of the evaluation index X _i in the data set R is obtained;

Step B, according to the entropy of the risk level in the data set R and the weighted value of the risk level entropy of the evaluation index _Xi in the data set R, the information gain of the evaluation index _Xi is obtained;

Step C, using the information gain of the evaluation index X _i and the entropy of the number of elements in all subsets corresponding to the evaluation index _Xi relative to the number of elements in the data set R, to calculate the information gain rate of the evaluation index X _i ;

After obtaining the information gain ratios of the n evaluation indexes, the information gain ratios are sorted to obtain the importance levels of the n evaluation indexes X ₁ , X ₂ , . . . , Xi _, . . . , X _n .

Preferably, the method also has the following characteristics: the calculation expression adopted for calculating the entropy H(R) of the risk level in the data set R is:

$\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; per</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; *</span>{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; per</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span>$

Among them, C _j represents a risk level of the system, where j=1,...,K, per(C _j , S) means that the number of systems in the data set R belonging to the risk level C _j accounts for all of the total data set R The ratio of the number of systems.

Preferably, the method also has the following characteristics: the calculation method used to obtain the weighted value of the risk level entropy of the evaluation index X _i in the data set R is obtained by calculating the entropy of the risk level in each sub-set corresponding to the evaluation index X _i The expression is:

${\mathrm{<span\; class="notranslate">\; h</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; l</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\frac{<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; |</span>}{<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; |</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; )</span>$

Among them, l represents the type of scoring results of the evaluation index _Xi in the data set S;

|R| indicates the number of risk levels in the data set R;

|R _i | indicates the number of risk levels in the subset R _i ;

H(R _i ) represents the entropy of the risk levels in the set R _i .

Preferably, the method also has the following characteristics: the information gain of the evaluation index _Xi and the entropy of the number of elements in all subsets corresponding to the evaluation index _Xi relative to the number of elements in the data set R are used to calculate the evaluation index The information gain rate of _Xi , including:

The calculation expression for calculating the entropy of the number of elements in all subsets corresponding to the evaluation index _Xi relative to the number of elements in the data set R is

$\mathrm{<span\; class="notranslate">\; split</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; infox</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; l</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; (</span><span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; *</span>{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; ;</span>$

Calculate the ratio of the information gain of the evaluation index _Xi to the entropy of the number of elements in all subsets corresponding to the evaluation index _Xi to the number of elements in the data set R, and obtain the information gain rate of the evaluation index _Xi .

A method for correcting scoring results of evaluation indicators, comprising:

Using the above method, the importance level of the evaluation index is obtained;

The scoring results of the evaluation indicators currently made by the system are corrected by using the importance levels of the obtained evaluation indicators, and the corrected scoring results of each evaluation indicator are obtained.

An acquisition device for evaluating the importance level of indicators, comprising:

The save module is used to save the scoring results of m systems on n evaluation indicators X ₁ , X ₂ ,...,X _i ,...,X _n in the data set S, and save the risk of each system The ranks are saved in the dataset R;

A computing module, connected to the saving module, used to calculate the entropy H(R) of the risk level in the data set R;

The execution module is connected with the storage module and the calculation module, and is used to execute steps A to C for each evaluation indicator, including:

Step A. According to the scoring results of the evaluation index _Xi in the data set S, all the risk levels corresponding to the system with the same scoring results for the evaluation index _Xi in the data set R are used as the evaluation index _Xi in the data set R For a corresponding sub-set, by calculating the entropy of the risk level in each sub-set corresponding to the evaluation index X _i , the weighted value of the risk level entropy of the evaluation index X _i in the data set R is obtained;

Step B, according to the entropy of the risk level in the data set R and the weighted value of the risk level entropy of the evaluation index _Xi in the data set R, the information gain of the evaluation index _Xi is obtained;

Step C, using the information gain of the evaluation index X _i and the entropy of the number of elements in all subsets corresponding to the evaluation index _Xi relative to the number of elements in the data set R, to calculate the information gain rate of the evaluation index X _i ;

A sorting module, connected to the calculation module, used to sort the obtained information gain ratios after obtaining the information gain ratios of n evaluation indexes, and obtain n evaluation indexes X ₁ , X ₂ ,...,X _i ,... , the importance level of X _n .

Preferably, the device also has the following characteristics: the calculation expression adopted by the calculation module to calculate the entropy H(R) of the risk level in the data set R is:

$\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; per</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; *</span>{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; per</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span>$

Among them, C _j represents a risk level, where j=1,...,K, per(C _j , S) means that the number of systems in the data set R belonging to the risk level C _j accounts for the total number of systems in the data set R proportion.

Preferably, the device also has the following characteristics: the execution module obtains the weighted value of the risk level entropy of the evaluation index X _i in the data set R by calculating the entropy of the risk level in each subset corresponding to the evaluation index X _i The calculation expression used is:

${\mathrm{<span\; class="notranslate">\; h</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; l</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\frac{<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; |</span>}{<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; |</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; )</span>$

Among them, l represents the type of scoring results of the evaluation index _Xi in the data set S;

|R| indicates the number of risk levels in the data set R;

|R _i | indicates the number of risk levels in the subset R _i ;

H(R _i ) represents the entropy of the risk levels in the set R _i .

Preferably, the device also has the following characteristics:

The execution module is used to calculate the entropy of the number of elements in all subsets corresponding to the evaluation index Xi relative to the number of elements in the data set _R when calculating the information gain rate of the evaluation index X _i , and calculate the evaluation index The ratio of the information gain of _Xi to the entropy of the number of elements in all subsets corresponding to the evaluation index _Xi to the number of elements in the data set R is obtained to obtain the information gain rate of the evaluation index _Xi ;

Wherein the calculation expression of the entropy of the number of elements in all subsets corresponding to the evaluation index X _i relative to the number of elements in the data set R is:

$\mathrm{<span\; class="notranslate">\; split</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; infox</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; l</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; (</span><span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; *</span>{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; .</span>$

A correction system for scoring results of evaluation indicators, comprising:

obtaining means for obtaining the importance level of the evaluation index by using the above means;

The correction device is used for correcting the scoring results of the evaluation indicators currently made by the system by using the obtained importance levels of the evaluation indicators, and obtaining the corrected scoring results of each evaluation indicator.

In the device embodiment provided by the present invention, the importance of the influence of the evaluation index on the risk level is obtained by calculating the information gain rate of the evaluation index in the same industry. The information gain rate is the ratio of the number of elements in all subsets corresponding to the information gain and the evaluation index to the entropy of the number of elements in the data set R. Therefore, compared with using information gain as the basis for evaluating the importance of the index, based on information The method of gain rate is more reasonable, and it will not be biased towards indicators with more values.

Description of drawings

FIG. 1 is a schematic flow diagram of an embodiment of a method for obtaining the importance level of an evaluation indicator provided by the present invention;

FIG. 2 is a schematic structural diagram of an embodiment of a system for acquiring the importance level of evaluation indicators provided by the present invention.

Detailed ways

In order to make the purpose, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that, in the case of no conflict, the embodiments in the present application and the features in the embodiments can be combined arbitrarily with each other.

FIG. 1 is a schematic flowchart of an embodiment of a method for obtaining the importance level of an evaluation indicator provided by the present invention. The method embodiment shown in Fig. 1 comprises:

Step 101, save the scoring results made by m systems on n evaluation indicators X ₁ , X ₂ ,...,X _i ,...,X _n in the data set S, and save the risk level of each system saved in the dataset R;

Taking the securities industry as an example, according to GB/T 20984-2007 "Information Security Technology Information Security Risk Assessment Specification" and the characteristics of the securities industry, the assessment elements are further subdivided by analyzing the assets, threats, vulnerabilities and existing measures of the system . Assets are divided into asset confidentiality (X ₁ ), asset integrity (X ₂ ), asset availability (X ₃ ) for assignment; threats from nature (X ₄ ), environment (X ₅ ), unauthorized personnel (X ₆ ) and Authorized personnel (X ₇ ) conduct threat analysis; vulnerability identification from security management (X ₈ ), system operation and maintenance (X ₉ ), business application system (X ₁₀ ), network (X ₁₁ ), network equipment (X ₁₂ ), operating system (X ₁₃ ), database (X ₁₄ ), and security device vulnerabilities (X ₁₅ ); existing security measures include security management and security operation and maintenance (X ₁₆ ), business application security protection (X ₁₇ ), network Consider the five measures of security protection (X ₁₈ ), operating system security (X ₁₉ ) and database security protection (X ₂₀ ). So there are 20 evaluation metrics in total.

Table 1 shows the scores and risk levels of the evaluation indicators of 10 different systems in the securities industry. The scoring result of the indicator takes an integer between 1 and 9, and the greater the value, the greater the risk. The risk level of the system is 5 levels, ranked as 5, 4, 3, 2, 1 in descending order.

Table 1

Step 102, obtain the entropy of the data set R, and its calculation expression is:

$\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; per</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; *</span>{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; per</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span>$

Among them, C _j represents a risk level, where j=1, 2, ..., k, k≤5, which is recorded as {C ₁ , C ₂ , ..., C _k } in turn, per(C _j , S) represents the data The ratio of the number of systems belonging to the risk level C _j in the set R to the total number of systems in the total data set R;

The above example illustrates that according to the risk level, the risk levels of 10 systems can be divided into 5 categories, and the number of systems in each category is 2, so

$\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; per</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; *</span>{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; per</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>$

$<span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; 5</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\frac{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; 10</span>}}{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\frac{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; 10</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; 5</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 2.3219</span>}$

Step 103, calculate the information gain rate of each evaluation index, specifically:

Perform steps A to C for each evaluation indicator, including:

Step A. According to the scoring results of the evaluation index _Xi in the data set S, all the risk levels corresponding to the system with the same scoring results for the evaluation index _Xi in the data set R are used as the evaluation index _Xi in the data set R For a corresponding sub-set, by calculating the entropy of the risk level in each sub-set corresponding to the evaluation index X _i , the weighted value of the risk level entropy of the evaluation index X _i in the data set R is obtained; the specific calculation expression is:

${\mathrm{<span\; class="notranslate">\; h</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; l</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\frac{<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; |</span>}{<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; |</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; )</span>$

|R| represents the number of systems in the data set R

|R _i | represents the number of systems of the subset R _i

H(R _i ) represents the entropy of the risk levels in the subset R _i .

Take the calculation of the weighted value of the subset entropy of the risk level of the evaluation index _X1 as an example to illustrate:

First, the risk level of the evaluation index X ₁ is statistically calculated, specifically, the value of the evaluation index X ₁ in the 10 systems is statistically analyzed. Then count the number of risk level systems corresponding to each value of the index X ₁ and the corresponding system risk level. The specific statistical results are as follows: The evaluation index X ₁ has 6 values in 10 systems: 1, 2, 3, 5, 6, 7, so R={R ₁ , R ₂ ,...,R ₆ } ,in:

There are 3 systems with a value of 1, and the final risk levels of the 3 systems are 1, 3, 4 respectively, denoted as R ₁ ;

If the value is 2, there is 1 system, and the corresponding system risk level is 2, denoted as R ₂ ;

There are 3 systems with a value of 3, and the final risk levels of the 3 systems are 1, 3, and 5 respectively, which are denoted as R ₃ ;

The values of 5, 6, and 7 have one system respectively, and the corresponding system risk levels are 4, 5, and 2, which are respectively recorded as R ₄ , R ₅ , and R ₆ .

According to the above statistical results, the calculation results are as follows:

H(R ₁ )=H(R ₃ )=-0.1*3*log ₂ 0.1=0.9966

H(R ₂ )=H(R ₄ )=H(R ₅ )=H(R ₆ )=-0.1*log ₂ 0.1=0.3322

${\mathrm{<span\; class="notranslate">\; h</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; l</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\frac{<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; |</span>}{<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; |</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; )</span>$

$<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0.3</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 0.1</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 0.3</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 4</span>}}{\overset{\mathrm{<span\; class="notranslate">\; 6</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; 0.1</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>$

$<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0.7308</span>}$

Step B, according to the entropy of the risk level in the data set R and the weighted value of the risk level entropy of the evaluation index _Xi in the data set R, the information gain of the evaluation index _Xi is obtained;

$\mathrm{<span\; class="notranslate">\; Gain</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span>$

Taking the calculation of the information gain of the evaluation index _X1 as an example, we can get:

$\mathrm{<span\; class="notranslate">\; Gain</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; h</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1.5911</span>}$

Step C, using the information gain of the evaluation index _Xi and the entropy of the number of elements in all subsets corresponding to the evaluation index _Xi relative to the number of elements in the data set R, to calculate the information gain rate of the evaluation index _Xi ; specifically include:

Calculate the entropy of the number of elements in all subsets corresponding to the evaluation index _Xi relative to the number of elements in the data set R. The calculation expression is

$\mathrm{<span\; class="notranslate">\; split</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; infox</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; j</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; (</span><span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; *</span>{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>$

For the evaluation index _X1 ,

$\mathrm{<span\; class="notranslate">\; split</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; infox</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; the\; s</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; (</span><span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; log</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>$

$<span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; 0.3</span>}<span\; class="notranslate">\; *</span>{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; 0.3</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; 0.1</span>}<span\; class="notranslate">\; *</span>{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; 0.1</span>}$

$<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 2.3710</span>}$

The calculation expression for calculating the information gain rate of index _Xi is:

$\mathrm{<span\; class="notranslate">\; Gain\; ratio</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; Gain</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; split</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; infox</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>}$

Therefore, the information gain rate of the evaluation index _X1 is:

$\mathrm{<span\; class="notranslate">\; Gain\; ratio</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; Gain</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; split</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; infox</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0.6711</span>}$

According to the above processing flow, the information gain rate of other indicators can also be obtained:

Step 104: Sort the obtained information gain ratios of the evaluation indicators to obtain the importance levels of the evaluation indicators.

The importance of the evaluation elements is ranked in descending order: network device vulnerability (X ₁₂ ), database vulnerability (X ₁₄ ), security device vulnerability (X ₁₅ ), environment (X ₅ ) threat, asset availability (X ₃ ), network security protection measures (X ₁₈ ), business application security protection measures (X ₁₇ ), natural (X ₄ ) threats, asset integrity (X ₂ ), system operation and maintenance vulnerabilities (X ₉ ), threats to authorized personnel (X ₇ ), security management and security operation and maintenance measures (X ₁₆ ), security management vulnerability (X ₈ ), operating system security (X ₁₉ ) measures, network (X ₁₁ ) vulnerability, operating system vulnerability (X ₁₃ ), business application system vulnerability (X ₁₀ ), asset confidentiality (X ₁ ), database security protection measures (X ₂₀ ), and threat of unauthorized personnel (X ₆ ).

According to the analysis of the data based on the results of the risk assessment, it is concluded that in order to ensure the security of the system, the five security elements that should be given priority in the securities industry are the vulnerability of network equipment, the vulnerability of databases, the vulnerability of security equipment, and environmental threats. And asset availability, these five key evaluation indicators are relatively consistent with the content of the five security entities currently considered by the securities industry. The system of the securities industry can formulate corresponding rectification measures in terms of security rectification according to the importance of evaluation indicators and the current system situation.

It should be noted that this embodiment takes the importance level of each evaluation index in the securities industry as an example for illustration, but it is not limited to this, and the importance level of evaluation indicators in other industries can also be calculated, as long as the input is different in the industry. The system can obtain the importance index of the evaluation index of information security in the industry by scoring the result composed of the same evaluation index; similarly, this embodiment deals with the evaluation index in the field of information security, but it is not limited to this, and can also The importance of obtaining evaluation indicators for the industry's internal transaction processing capabilities.

In addition, the reason why this embodiment can objectively and accurately calculate the importance level of the evaluation index is because the calculation expression in step 103 is used to calculate the information gain rate of the evaluation index in the same industry, and the evaluation index has a significant impact on the evaluation index. The importance of the impact of the risk level. The information gain rate is the ratio of the information gain to the entropy of the sub-sets obtained by dividing the index value. Therefore, compared with using the information gain as the basis for evaluating the importance of the index, the method based on the information gain rate is more reasonable and will not bias the index selection. A multi-value indicator.

With reference to the above analysis, it can be seen that the calculation expression of the information gain rate used in step 103 is used in machine learning to establish a data classification model when classifying data, which is completely different from the method usually used by those skilled in the art, and opens up the calculation expression a new field of application.

In the past, the ranking of the importance of indicators was obtained through the subjective judgment of experts on the comparison of indicators, which was highly subjective. The present invention ranks the importance of indexes obtained by calculating the information gain rate of the indexes as a whole, thus more accurately and objectively reflecting the importance of indexes in the entire evaluation index system. The ranking of the importance of the indicators can correct the ranking of the importance of the indicators by the experts, thereby making the risk level more accurate.

The importance of evaluation indicators in the same industry is obtained to provide guidance for the rectification work of information system security construction in this industry. In order to make the rectification work more targeted, it is necessary to focus on the rectification of the evaluation indicators that have an important impact on the risk level, which is very necessary to improve the efficiency of the rectification work.

In addition, the present invention also provides an embodiment of a correction method for scoring results of evaluation indicators, including:

Using the method described above, the importance level of the evaluation index is obtained;

Using the importance level of the evaluation index obtained, the scoring result of the evaluation index currently made by the system is corrected to obtain the corrected scoring result of each evaluation index.

For example, if there are two evaluation indicators A and B, the scoring result of the existing evaluation indicator A is lower than that of the evaluation indicator B, but according to the obtained importance level, the evaluation indicator A is higher than the evaluation indicator B. If the importance is high, the scoring result of at least one of the evaluation indicators A and B is corrected so that the scoring result of the evaluation indicator A is higher than the scoring result of the evaluation indicator B; and vice versa.

In the method embodiment provided by the present invention, the scoring results of multiple systems for the same evaluation index are counted, and based on the above statistical results, the information gain rate of each evaluation index is calculated, and then the importance level of the evaluation index is obtained according to the size of the information gain rate , providing a theoretical basis for objectively and accurately setting the scoring results of each evaluation index, which is more accurate and objective compared with the setting based on human experience in the prior art.

FIG. 2 is a schematic structural diagram of a device for acquiring importance levels of evaluation indicators provided by the present invention. The device shown in Figure 2 includes:

The saving module 201 is used to save the scoring results made by m systems on n evaluation indicators X ₁ , X ₂ ,...,X _i ,...,X _n in the data set S, and save each system's The risk level is saved in the dataset R;

Calculation module 202, connected to the preservation module 201, for calculating the entropy H(R) of the risk level in the data set R;

The execution module 203 is connected to the saving module 201 and the calculation module 202, and is used to execute steps A to C for each evaluation index, including:

Step A. According to the scoring results of the evaluation index _Xi in the data set S, all the risk levels corresponding to the system with the same scoring results for the evaluation index _Xi in the data set R are used as the evaluation index _Xi in the data set R For a corresponding sub-set, by calculating the entropy of the risk level in each sub-set corresponding to the evaluation index X _i , the weighted value of the risk level entropy of the evaluation index X _i in the data set R is obtained;

Step B, according to the entropy of the risk level in the data set R and the weighted value of the risk level entropy of the evaluation index _Xi in the data set R, the information gain of the evaluation index _Xi is obtained;

Step C, using the information gain of the evaluation index X _i and the entropy of the number of elements in all subsets corresponding to the evaluation index _Xi relative to the number of elements in the data set R, to calculate the information gain rate of the evaluation index X _i ;

The sorting module 204 is connected to the calculation module 203, and is used to sort the obtained information gain ratios after obtaining the information gain ratios of n evaluation indexes, so as to obtain n evaluation indexes X ₁ , X ₂ ,...,X _i , ..., the importance level of X _n .

Wherein, the calculation expression adopted by the calculation module 202 in calculating the entropy H(R) of the risk level in the data set R is:

$\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; per</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; *</span>{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; per</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span>$

Among them, C _j represents a risk level, where j=1,...,K, per(C _j , S) means that the number of systems in the data set R belonging to the risk level C _j accounts for the total number of systems in the data set R proportion.

Wherein, the execution module 203 obtains the weighted value of the risk level entropy of the evaluation index X _i in the data set R by calculating the entropy of the risk level in each sub-set corresponding to the evaluation index X _i . The calculation expression used is:

${\mathrm{<span\; class="notranslate">\; h</span>}}_{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; l</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\frac{<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; |</span>}{<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; |</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; )</span>$

Among them, l represents the type of scoring results of the evaluation index _Xi in the data set S;

|R| indicates the number of risk levels in the data set R;

|R _i | indicates the number of risk levels in the subset R _i ;

H(R _i ) represents the entropy of the risk levels in the set R _i .

Wherein, the execution module 203 is configured to calculate the entropy of the number of elements in all subsets corresponding to the evaluation index _Xi relative to the number of elements in the data set R when calculating the information gain rate of the evaluation index _Xi , and pass Calculate the ratio of the information gain of the evaluation index X _i to the entropy of the number of elements in all subsets corresponding to the evaluation index X _i relative to the number of elements in the data set R, and obtain the information gain rate of the evaluation index X _i ;

Wherein the calculation expression of the entropy of the number of elements in all subsets corresponding to the evaluation index X _i relative to the number of elements in the data set R is:

$\mathrm{<span\; class="notranslate">\; split</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; infox</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; l</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; (</span><span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; *</span>{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; .</span>$

In the device embodiment provided by the present invention, the importance of the influence of the evaluation index on the risk level is obtained by calculating the information gain rate of the evaluation index in the same industry. The information gain rate is the ratio of the number of elements in all subsets corresponding to the information gain and the evaluation index to the entropy of the number of elements in the data set R. Therefore, compared with using information gain as the basis for evaluating the importance of the index, based on information The method of gain rate is more reasonable, and it will not be biased towards indicators with more values.

In addition, the present invention provides a correction system for scoring results of evaluation indicators, including:

The obtaining device is used to obtain the importance level of the evaluation index by using the device shown in Figure 2;

The correction device is used for correcting the scoring results of the evaluation indicators currently made by the system by using the obtained importance levels of the evaluation indicators, and obtaining the corrected scoring results of each evaluation indicator.

The system embodiment provided by the present invention counts the scoring results of multiple systems for the same evaluation index, and calculates the information gain rate of each evaluation index based on the above statistical results, and then obtains the importance level of the evaluation index according to the information gain rate , which provides a theoretical basis for objectively and accurately setting the scoring results of each evaluation index, which is more accurate and objective compared with the setting based on human experience in the prior art.

The above is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Anyone skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present invention. Should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope described in the claims.

Claims (10)

1. other acquisition methods of the significance level of an evaluation index is characterized in that, comprising:

To obtain m system respectively to n evaluation index X ₁, X ₂..., X _i..., X _nThe appraisal result of having done is kept among the data set S, and the risk class of each system is kept among the data set R;

The entropy H (R) of risk class among the computational data collection R;

To each evaluation index execution in step A～C, comprising:

Steps A, according to evaluation index X among the data set S _iAppraisal result, will be to evaluation index X _iThe system that makes identical appraisal result at the pairing whole risk class of data set R as evaluation index X _iA pairing son set in data set R is through calculating evaluation index X _iThe entropy of risk class obtains evaluation index X among the data set R in each corresponding subclass _iThe weighted value of risk class entropy;

Step B, according to evaluation index X among the entropy of risk class among the data set R and the data set R _iThe weighted value of risk class entropy, obtain evaluation index X _iInformation gain;

Step C, employing evaluation index X _iInformation gain and evaluation index X _iElement number calculates evaluation index X with respect to the entropy of element number among the data set R in corresponding whole subclass _iThe information gain rate;

After obtaining the information gain rate of n evaluation index, the information gain rate is sorted, thereby obtain n evaluation index X ₁, X ₂..., X _i..., X _nImportance information.

2. method according to claim 1 is characterized in that, the calculation expression that the entropy H (R) of risk class is adopted among the said computational data collection R is:

$H\left(R\right)=-\underset{j=1}{\overset{k}{\mathrm{\&Sigma;}}}\mathrm{per}(C_j,R)*{\log }_2\mathrm{per}(C_j,R)$

Wherein, C _jA kind of risk class of expression system, j=1 wherein ..., K, per (C _j, S) belonging to risk class among the expression data set R is C _jSystem's number account among the total data collection R all ratios of system quantities.

3. method according to claim 1 is characterized in that, and is said through calculating evaluation index X _iThe entropy of risk class obtains evaluation index X among the data set R in each corresponding subclass _iThe calculation expression that weighted value adopted of risk class entropy be:

$H_{X_i}\left(R\right)=\underset{j=1}{\overset{l}{\mathrm{\&Sigma;}}}\frac{\left|R_j\right|}{\left|R\right|}*H\left(R_j\right)$

Wherein, l representes evaluation index X among the data set S _iThe kind of appraisal result;

| R| representes the number of risk class among the data set R;

| R _i| expression subclass R _iThe number of middle risk class;

H (R _i) expression set R _iThe entropy of middle risk class.

4. method according to claim 3 is characterized in that, said employing evaluation index X _iInformation gain and evaluation index X _iElement number calculates evaluation index X with respect to the entropy of element number among the data set R in corresponding whole subclass _iThe information gain rate, comprising:

Calculate evaluation index X _iElement number with respect to the calculation expression of the entropy of element number among the data set R does in corresponding whole subclass

$\mathrm{Split}\_\mathrm{Infox}\left(X_i\right)=-\underset{i=1}{\overset{l}{\mathrm{\&Sigma;}}}(\left(\right|R_i|/|R\left|\right)*{\log }_2\left(\right|R_i|/|R\left|\right));$

Calculate evaluation index X _iInformation gain and evaluation index X _iElement number obtains evaluation index X with respect to the ratio of the entropy of element number among the data set R in corresponding whole subclass _iThe information gain rate.

5. the modification method of the appraisal result of an evaluation index is characterized in that, comprising:

Adopt like the arbitrary described method of claim 1 to 4, obtain the importance information of evaluation index;

Employing obtains the importance information of evaluation index the appraisal result of the current evaluation index of making of system is revised, and obtains the appraisal result of revised each evaluation index.

6. other deriving means of the significance level of an evaluation index is characterized in that, comprising:

Preserve module, be used for obtaining m system respectively to n evaluation index X ₁, X ₂..., X _i..., X _nThe appraisal result of having done is kept among the data set S, and the risk class of each system is kept among the data set R;

Computing module links to each other with said preservation module, is used for the entropy H (R) of computational data collection R risk class;

Execution module links to each other with said computing module with said preservation module, is used for each evaluation index execution in step A～C is comprised:

Steps A, according to evaluation index X among the data set S _iAppraisal result, will be to evaluation index X _iThe system that makes identical appraisal result at the pairing whole risk class of data set R as evaluation index X _iA pairing son set in data set R is through calculating evaluation index X _iThe entropy of risk class obtains evaluation index X among the data set R in each corresponding subclass _iThe weighted value of risk class entropy;

Step B, according to evaluation index X among the entropy of risk class among the data set R and the data set R _iThe weighted value of risk class entropy, obtain evaluation index X _iInformation gain;

Step C, employing evaluation index X _iInformation gain and evaluation index X _iElement number calculates evaluation index X with respect to the entropy of element number among the data set R in corresponding whole subclass _iThe information gain rate;

Order module links to each other with said computing module, is used for after obtaining the information gain rate of n evaluation index, the information gain rate that obtains being sorted, and obtains n evaluation index X ₁, X ₂..., X _i..., X _nImportance information.

7. device according to claim 6 is characterized in that, the calculation expression that the entropy H (R) of said computing module risk class in computational data collection R is adopted is:

$H\left(R\right)=-\underset{j=1}{\overset{k}{\mathrm{\&Sigma;}}}\mathrm{per}(C_j,R)*{\log }_2\mathrm{per}(C_j,R)$

Wherein, C _jRepresent a kind of risk class, j=1 wherein ..., K, per (C _j, S) belonging to risk class among the expression data set R is C _jSystem's number account among the total data collection R all ratios of system quantities.

8. device according to claim 6 is characterized in that, said execution module is passing through to calculate evaluation index X _iThe entropy of risk class obtains evaluation index X among the data set R in each corresponding subclass _iThe calculation expression that weighted value adopted of risk class entropy be:

$H_{X_i}\left(R\right)=\underset{j=1}{\overset{l}{\mathrm{\&Sigma;}}}\frac{\left|R_j\right|}{\left|R\right|}*H\left(R_j\right)$

Wherein, l representes evaluation index X among the data set S _iThe kind of appraisal result;

| R| representes the number of risk class among the data set R;

| R _i| expression subclass R _iThe number of middle risk class;

H (R _i) expression set R _iThe entropy of middle risk class.

9. device according to claim 8 is characterized in that:

Said execution module is used for calculating evaluation index X _iThe information gain rate time, calculate evaluation index X _iElement number is with respect to the entropy of element number among the data set R in corresponding whole subclass, and through calculating evaluation index X _iInformation gain and evaluation index X _iElement number obtains evaluation index X with respect to the ratio of the entropy of element number among the data set R in corresponding whole subclass _iThe information gain rate;

Wherein said evaluation index X _iElement number with respect to the calculation expression of the entropy of element number among the data set R does in corresponding whole subclass

$\mathrm{Split}\_\mathrm{Infox}\left(X_i\right)=-\underset{i=1}{\overset{l}{\mathrm{\&Sigma;}}}(\left(\right|R_i|/|R\left|\right)*{\log }_2\left(\right|R_i|/|R\left|\right)).$

10. the update the system of the appraisal result of an evaluation index is characterized in that, comprising:

Deriving means is used for adopting like the arbitrary described device of claim 6 to 9, obtains the importance information of evaluation index;

Correcting device is used to adopt the importance information that obtains evaluation index that the appraisal result of the current evaluation index of making of system is revised, and obtains the appraisal result of revised each evaluation index.

Images