CN106411854B - A kind of network security risk evaluation method based on fuzzy Bayes - Google Patents
A kind of network security risk evaluation method based on fuzzy Bayes Download PDFInfo
- Publication number
- CN106411854B CN106411854B CN201610805770.2A CN201610805770A CN106411854B CN 106411854 B CN106411854 B CN 106411854B CN 201610805770 A CN201610805770 A CN 201610805770A CN 106411854 B CN106411854 B CN 106411854B
- Authority
- CN
- China
- Prior art keywords
- risk
- reliability
- value
- risk class
- fuzzy
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L63/00—Network architectures or network communication protocols for network security
- H04L63/20—Network architectures or network communication protocols for network security for managing network security; network security policies in general
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Computer Security & Cryptography (AREA)
- Computer Hardware Design (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of network security risk evaluation methods based on fuzzy Bayes, the present invention proposes a kind of opinion rating reliability algorithm according to the thought of theorem in Euclid space vector projection, the algorithm can integrate multidigit expertise and can handle the case where expert provides multiple evaluation results due to uncertainty, then Fuzzy processing is carried out to evaluation result by Gauss subordinating degree function again, finally solves the risk size that information measured system is faced in conjunction with the reasoning algorithm of Bayesian network model.The method can enhance the objectivity and validity of assessment result, so that the control and management for subsequent risk provide more reasonable effective foundation.
Description
Technical field
The present invention relates to a kind of safety risk estimating method, specially a kind of mould decomposed based on reliability vector rectangular projection
Bayesian network safety risk estimating method is pasted, network technique field is belonged to.
Background technique
With the development of internet technology with the continuous innovation of cyber-attack techniques, many safety problems of network bring are
As focus concerned by people, one of the important content that information security risk evaluation works as security assurance information is only most
Objective effectively risk may be identified and be quantified, just can guarantee subsequent prevention and neutralizing to risk, controlled
Within the scope of acceptable.
Common Information Security Risk Assessment Methods can substantially be divided into three classes: quantitative evaluating method, qualitative evaluation method
And comprehensive estimation method.Typical quantitative evaluating method mainly has clustering methodology, factor analysis, equal risk curve etc., this
Class method can more objectively indicate assessment result with data, but the data quantified sometimes may twist problem；It is typical
Qualitative evaluation method mainly have Application of Delphi Method, analysis, comparative etc., such method can be more fully deep
Ground reflects assessment result, but there is stronger artificial subjectivity sometimes；Typical comprehensive estimation method mainly has levels analysis
Method etc., such method, which will be assessed qualitatively and quantitatively, to be combined, and makes full use of the knowledge and experience of expert again with quantization as far as possible
Method keep assessment result more objective credible, however how the two effectively to be combined according to particular problem, is given full play to quantitative
It is the direction that people constantly endeavour with qualitative respective advantage.
It is proposed currently, researchers are based on fuzzy set, D-S evidence theory, gray theory and machine learning scheduling theory
A series of improved methods of risk assessment, advance the progress of information security risk evaluation.Fuzzy synthetic appraisement method is
Analysis method that is a kind of qualitative and quantitatively combining, converts quantitative assessment for qualitative evaluation by being subordinate to topology degree, can be effective
It handles expert and is difficult to the case where quantifying in the subjectivity of evaluation process and objective blooming；D-S evidence theory is a kind of not true
Qualitative inference method is capable of handling the uncertain problem due to caused by not knowing and is inaccurate；The research pair of gray theory
As being mainly information content deficiency, the system with imperfection, a series of assessment prediction work can be carried out to object；Engineering
The process for practising analog mankind study constructs Rule new knowledge according to existing knowledge to carry out risk assessment.
During information security risk evaluation, if thering is multidigit expert to evaluate each venture influence factor, simultaneously
Multiple evaluation results are provided due to the uncertainty of itself again, the diversity and probabilistic feelings for this assessment result
Condition, existing method tend not to well carry out it integrated treatment, and the objectivity of acquired results is to be improved.Therefore, it studies
A kind of evaluation result that can comprehensively consider multidigit expert and its uncertain situation are of great practical significance.
Summary of the invention
The object of the invention is that provide a kind of based on reliability vector rectangular projection decomposition to solve the above-mentioned problems
Fuzzy Bayesian network safety risk estimating method, to handle the uncertain factor due to expert and provide it is multiple evaluation etc.
The case where grade, while improving the objectivity and validity of assessment result.
The present invention is achieved through the following technical solutions above-mentioned purpose, a kind of network security risk based on fuzzy Bayes
Appraisal procedure, this method comprises:
S101, the evaluation result that multidigit expert is integrated by reliability vector rectangular projection decomposition algorithm, calculate each risk shadow
The factor of sound is in the reliability of different risk class；
S102, the reliability for being in different risk class to each venture influence factor calculated in S101 quantify, then benefit
Fuzzy processing is carried out to the opinion rating quantized value that expert provides with Gauss subordinating degree function, calculating is under the jurisdiction of different risks etc.
The degree of grade, is weighted summation in conjunction with reliability, calculates the probability value that each venture influence factor is in different risk class；
S103, the probability value that each venture influence factor calculated in S102 is in different risk class is input to by pattra leaves
In the information security risk evaluation index system of this network struction, the probability that entire information system is in different risk class is calculated
Value；
S104, basis " gravity model appoach " carry out anti fuzzy method processing, quantify the risk size of entire information system.
Preferably, the evaluation result by the comprehensive multidigit expert of reliability vector rectangular projection decomposition algorithm in S101, meter
The reliability that each venture influence factor is in different risk class is calculated, is specifically included:
1) opinion rating result of the K experts to a certain venture influence factor H, is denoted as Vi(wherein i=1,2 ..., K),
Again simultaneously by all evaluation results phases, it is indicated, is denoted as with identification framework ΩBecause of Ω
In N number of element do not include mutually two-by-two, make following analogy using the thought of theorem in Euclid space:
(1) identification framework Ω is regarded as the N-dimensional vector space comprising N number of reference axis；
(2) the N number of element not included mutually in Ω is regarded as orthogonal reference axis, i.e. x two-by-two1,x2,…,xj,…,
xn；
(3) by the opinion rating result V of every expertiRegard a reliability vector v of N-dimensional vector space Ω asi(i=1,
2,…,K)；
If 2), certain opinion rating result ViThere is M element, enables ViCorresponding reliability vector viWith the angle of this M reference axis
It is equal, it is 90 degree with reference axis angles other in coordinate system, therefore, as reference axis xjIt is contained in reliability vector viWhen middle, viWith
The angular separation cosine of this reference axisAs reference axis xjIt is not included in reliability vector viWhen middle, viWith
The angular separation cosine of this reference axis
3) each reliability vector v, is calculatediMould | | vi| |, i.e., after the accuracy rate normalization of every expert as a result, then counting
Calculate each reliability vector viIn each reference axis xjRectangular projection decomposition value on (wherein j=1,2 ..., n), i.e.,
4), by each reliability vector viThe cumulative summation of rectangular projection decomposition value in the same reference axis, then returns again
One changes the reliability to get this venture influence factor H in each opinion rating.
Preferably, different risk class locating for each venture influence factor are quantified in S102, recycles Gauss
Subordinating degree function carries out Fuzzy processing to the opinion rating quantized value that expert provides, and calculates the journey for being under the jurisdiction of different risk class
Degree, specifically includes:
1) the risk size of a certain venture influence factor, is divided into N number of grade, quantized value is defined on [0,1] section, wind
Danger is bigger, and value is bigger, conversely, quantized value is smaller；
2) N number of Gauss subordinating degree function, is constructed according to N number of risk class
Wherein, the μ represents the center of subordinating degree function, and the selection of central value is determined according to the risk class of division
It is fixed, and in [0,1] section, it is uniformly distributed as far as possible；The σ indicates the width of subordinating degree function, reflects expert to certainly
The uncertainty for the evaluation result that oneself provides, σ is bigger, shows that expert is lower to the certainty factor of assessed value；
3) Fuzzy processing, is carried out to the opinion rating quantized value that expert provides by Gauss subordinating degree function, i.e., amount
Opinion rating after change is separately input in N number of subordinating degree function, renormalization, can acquire expert to this venture influence because
The opinion rating of element is under the jurisdiction of the degree of different risk class.
Preferably, the combination degree of membership in S102 and reliability are weighted summation, calculate each venture influence factor and are in not
With the probability value of risk class, specifically include: the reliability that reliability vector rectangular projection decomposition algorithm is acquired and is commented as weight
The degree that valence grade is under the jurisdiction of different risk class is multiplied, then the probability value of same risk class is added to get each risk shadow
The factor of sound is in the probability value of different risk class.
Preferably, the probability value that each venture influence factor is in different risk class in S103 is input to by Bayes
In the information security risk evaluation index system of network struction, the probability that entire information system is in different risk class is calculated
Value, specifically includes:
1), according to information security risk evaluation the relevant technologies and administrative standard, Bayesian network is constructed in conjunction with actual conditions
Network model；
Wherein, Bayesian network model is made of model structure and model parameter two parts, and model structure is one oriented
Acyclic figure is formed by representing the node of variable and representing causal directed arc between variable, and model parameter is then to represent to become
The conditional probability table CPT of relationship between amount；
2), Bayesian network model reasoning process specifically includes: if there is n concealed nodes H in a Bayesian network
={ H1,H2,…,HnAnd m Observable node O={ O1,O2,…,Om, node HiFather node be denoted as F (Hi), node Oj's
Father node is denoted as F (Oj), according to conditional independence assumption and d- law of segregation, the joint probability distribution of all variables isIt, can basis in conjunction with Bayesian formula
The probability of the probabilistic inference concealed nodes of observer nodes
3), using each venture influence factor as the Observable node of Bayesian network model, the risk of entire information system
As the root node of Bayesian network, each venture influence factor is in the Bayes that different grades of probability value is input to building
In network model, the probability value that entire information system is in different risk class is calculated.
It preferably, can be with when it is implemented, the conditional probability table CPT in Bayesian network model is got by expertise
It is tested repeatedly according to a large amount of sample data, appropriate adjustment is carried out to data in table, to improve the visitor of risk evaluation result
The property seen.
Preferably, anti fuzzy method processing is carried out according to " gravity model appoach ", quantifies the risk size of entire information system, it is specific to wrap
It includes: setting the center-of-gravity value of each risk class, be multiplied with corresponding probability value, then be added, just obtain entire information measured system
The faced value-at-risk of system.
The invention has the benefit that the present invention proposes a kind of opinion rating according to the thought of theorem in Euclid space vector projection
Reliability vector rectangular projection decomposition algorithm, the algorithm can integrate multidigit expertise and expert can be handled due to uncertainty and
Then the case where providing multiple evaluation results carries out Fuzzy processing to evaluation result by Gauss subordinating degree function again, finally
The risk size faced in conjunction with the reasoning algorithm solution information measured system of Bayesian network model.Compared with prior art,
The present invention overcomes too relying on the shortcomings that expert's subjectivity judges in traditional Bayesian network methods of risk assessment, particularly with
Expert's assessment result multiplicity and uncertain phenomenon can carry out reasonable data processing to it, to enhance assessment result
Objectivity and validity provide more reasonable effective foundation for the control and management of subsequent risk.
Detailed description of the invention
Fig. 1 is a kind of fuzzy Bayesian network safety wind decomposed based on reliability vector rectangular projection in the embodiment of the present invention
The flow diagram of dangerous appraisal procedure；
Fig. 2 is the flow diagram of opinion rating reliability vector rectangular projection decomposition algorithm in the embodiment of the present invention；
Fig. 3 is the information security risk evaluation index system based on Bayesian network in the embodiment of the present invention；
Fig. 4 is expert in the embodiment of the present invention to the evaluation result of each venture influence factor.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
The present invention calculates the risk size that information system is faced using the reasoning algorithm of Bayesian network, basic herein
On, a kind of opinion rating reliability vector rectangular projection decomposition algorithm, the calculation are proposed according to the thought of theorem in Euclid space vector projection
Method can integrate multidigit expertise and can handle the case where expert provides multiple evaluation results due to uncertainty.With it is existing
Technology is compared, and the present invention overcomes lacking for expert's subjectivity judgement is too relied in traditional Bayesian network methods of risk assessment
Point can carry out reasonable data processing to it, to enhance particularly with expert's assessment result multiplicity and uncertain phenomenon
The objectivity and validity of assessment result.For a better understanding of the present invention, below with a specific example to the present invention into
Row detailed description.
System embodiment
According to the relevant technologies and administrative standard, during information security risk evaluation, three fundamentals are related generally to,
It is assets A respectively, threatens T and fragility V.Three security attributes for evaluating assets are confidentiality A1, integrality A2 and availability
A3；The factor to threaten can be divided into environmental factor T1 and human factor T2；To fragility identified mainly from technology V1 and
It is carried out in terms of management V2 two.On this basis, the Information Security Risk in the embodiment of the present invention based on Bayesian network is constructed
Evaluation index system, referring to Fig. 3, this Bayesian network model includes 7 Observable nodes and 4 concealed nodes.Please five specially
Family evaluates 7 Observable venture influence factors, opinion rating be divided into very low (VL), low (L), medium (M), high (H) and
Very high (VH), referring to fig. 4, the accuracy rate of five experts is respectively 0.8,0.85,0.9,0.8,0.9 to evaluation result.
The embodiment of the invention provides a kind of fuzzy Bayesian network safety winds decomposed based on reliability vector rectangular projection
Dangerous appraisal procedure, referring to Fig. 1, this method comprises:
S101, the evaluation result that multidigit expert is integrated by reliability vector rectangular projection decomposition algorithm, calculate each risk shadow
The factor of sound is in the reliability of different risk class；
S102, different risk class locating for each venture influence factor are quantified, recycles Gauss subordinating degree function
Fuzzy processing is carried out to the opinion rating quantized value that expert provides, calculates the degree for being under the jurisdiction of different risk class, in conjunction with
Reliability is weighted summation, calculates the probability value that each venture influence factor is in different risk class；
S103, the probability value that each venture influence factor is in different risk class is input to by Bayesian network building
In information security risk evaluation index system, the probability value that entire information system is in different risk class is calculated；
S104, basis " gravity model appoach " carry out anti fuzzy method processing, quantify the risk size of entire information system.
Referring to fig. 2, which can for reliability vector rectangular projection decomposition algorithm flow diagram described in the embodiment of the present invention
The evaluation result of comprehensive multidigit expert, calculates the reliability that each venture influence factor is in different risk class, specifically includes:
Opinion rating result of five experts to assets confidentiality is denoted as Vi(wherein i=1,2 ..., 5), V1={ high },
V2={ medium }, V3={ medium, high }, V4={ medium }, V5={ high }, then simultaneously by all evaluation results phases, with identification framework Ω
It indicates, is denoted asBecause two elements in Ω do not include mutually, the thought of theorem in Euclid space is utilized
Make following analogy:
(1) identification framework Ω is regarded as the two dimensional vector space comprising two reference axis；
(2) two elements not included mutually in Ω are regarded as orthogonal reference axis, i.e. x1,x2, wherein x1It represents
" medium ", x2Represent "high"；
(3) by the opinion rating result V of every expertiRegard a reliability vector v of two dimensional vector space Ω asi(i=1,
2,…,5)；
V1There are an element "high", corresponding reliability vector v1With two reference axis x1And x2Angular separation be respectively
Withx1It is not included in v1In, therefore, v1With reference axis x1Angular separation cosinex2It is contained in v1In, because
This, v1With reference axis x2Angular separation cosine
Similarly, reliability vector v2Direction cosines with two reference axis are respectivelyWithReliability to
Measure v3Direction cosines with two reference axis are respectivelyWithReliability vector v4It is sat with two
The direction cosines of parameter are respectivelyWithReliability vector v5Distinguish with the direction cosines of two reference axis
It isWith
Calculate each reliability vector viMould | | vi| |, i.e., it is after the accuracy rate normalization of every expert as a result, | | v1| |=
0.8/ (0.8+0.85+0.9+0.8+0.9)=0.1882, | | v2| |=0.2, | | v3| |=0.2118, | | v4| |=0.1882, |
|v5| |=0.2118；
Calculate each reliability vector viIn each reference axis xjRectangular projection decomposition value on (wherein j=1,2 ..., n), i.e.,Calculated result is as follows:
By each reliability vector viThe cumulative summation of rectangular projection decomposition value in the same reference axis, obtains in reference axis
x1On the sum of projection value r1=0+0.2+0.1498+0.1882+0=0.538, in reference axis x2On the sum of projection value r2=
0.1882+0+0.1498+0+0.2118=0.5498 is obtained after normalization Up to assets guarantor
Close property makes risk be in medium reliability 0.4946, is 0.5054 in high reliability；
Similarly, assets integrality makes risk be in low reliability 0.4, is 0.4118 in medium reliability, in height
Reliability be 0.1882；Asset availability makes risk be in medium reliability 0.4, is 0.6 in high reliability；Environmental factor
So that risk is in very low reliability 0.3882, is 0.6118 in low reliability；Human factor makes risk be in medium letter
Degree is 0.5164, is 0.4836 in high reliability；Technology fragility makes risk be in very low reliability 0.2, low
Reliability is 0.8；Management fragility makes risk be in low reliability 0.6118, is 0.3882 in medium reliability.
Different risk class locating for each venture influence factor are quantified described in the embodiment of the present invention, are recycled high
This subordinating degree function carries out Fuzzy processing to the opinion rating quantized value that expert provides, and calculating is under the jurisdiction of different risk class
Degree specifically includes:
Five risk class are quantified, very low, low, medium, high and very high quantized value is set to 0.1 respectively, 0.3,
0.5,0.7 and 0.9；
Five Gauss subordinating degree functions are constructed according to five risk class
Wherein, the μ represents the center of subordinating degree function, and the selection of central value is determined according to the risk class of division
It is fixed, and in [0,1] section, it is uniformly distributed as far as possible；The σ indicates the width of subordinating degree function, reflects expert to certainly
The uncertainty for the evaluation result that oneself provides, σ is bigger, shows that expert is lower to the certainty factor of assessed value；If five in the present embodiment
The corresponding subordinating degree function of a different brackets is respectivelyEnable σ=0.1；
Fuzzy processing is carried out to the opinion rating quantized value that expert provides by Gauss subordinating degree function, if expert provides
Opinion rating be " very low ", quantized value 0.1, by 0.1 be input to five subordinating degree functions and normalize after, acquired results are
Being under the jurisdiction of the degree of five risk class for " very low " evaluation is HVL=(0.6511 0.3485 0.0004 0 0)；
Similarly, it is H that " low " evaluation, which is under the jurisdiction of the degree of five risk class,L=(0.0108 0.8576 0.1316 0
0)；The degree that " medium " evaluation is under the jurisdiction of five risk class is HM=(0 0.0404 0.9192 0.0404 0)；"high" is commented
The degree that valence is under the jurisdiction of five risk class is HH=(0 0 0.1315 0.8577 0.0108)；" very high " evaluation is under the jurisdiction of five
The degree of a risk class is HVH=(0 0 0.0004 0.3485 0.6511).
Combination degree of membership and reliability described in the embodiment of the present invention are weighted summation, calculate each venture influence factor and are in
The probability value of different risk class, specifically includes:
Assets confidentiality makes risk be in medium reliability 0.4946, is 0.5054 in high reliability, " medium " comments
The degree that valence is under the jurisdiction of five risk class is HM=(0 0.0404 0.9192 0.0404 0), "high" evaluation are under the jurisdiction of five
The degree of risk class is HH=(0 0 0.1315 0.8577 0.0108) calculate 0.4946 × HM+0.5054×HH=(0
0.0200 0.5211 0.4535 0.0054), i.e., assets confidentiality makes risk be in very low, low, medium, high and very five high
The probability of grade is respectively 0,0.0200,0.5211,0.4535,0.0054；
Similarly, assets integrality make risk be in the probability of five grades be respectively 0.0043,0.3597,0.4559,
0.1781,0.0020；Asset availability make risk be in the probability of five grades be respectively 0,0.0161,0.4466,0.5308,
0.0065；It is respectively 0.2594,0.6600,0.0806,0,0 that environmental factor, which makes risk be in the probability of five grades,；It is artificial because
It is respectively 0,0.0209,0.5383,0.4356,0.0052 that element, which makes risk be in the probability of five grades,；Technology fragility makes wind
Probability of the danger in five grades is respectively 0.1388,0.7558,0.1054,0,0；Management fragility makes risk be in five etc.
The probability of grade is respectively 0.0066,0.5404,0.4373,0.0157,0.
The probability value that each venture influence factor is in different risk class is input to by shellfish described in the embodiment of the present invention
In the information security risk evaluation index system of this network struction of leaf, calculates entire information system and be in the general of different risk class
Rate value, specifically includes:
Using each venture influence factor as the Observable node of Bayesian network model, the risk conduct of entire information system
Each venture influence factor is in the Bayesian network that different grades of probability value is input to building by the root node of Bayesian network
In model, calculate the probability value that entire information system is in different risk class, the results showed that risk be in it is very low, low, medium,
High and very high probability is respectively 0.0591,0.2208,0.5007,0.1709,0.0485；
When it is implemented, the conditional probability table CPT in Bayesian network model is got by expertise, it can be according to a large amount of
Sample data tested repeatedly, in table data carry out appropriate adjustment, to improve the objectivity of risk evaluation result.
Basis described in the embodiment of the present invention " gravity model appoach " carries out anti fuzzy method processing, quantifies the risk of entire information system
Size specifically includes:
The center-of-gravity value for setting this very low, low, medium, high and very high five risk class is respectively 0.1,0.3,0.5,0.7
With 0.9；
By upper gained, the corresponding probability of five risk class is respectively 0.0591,0.2208,0.5007,0.1709,
0.0485, calculate value-at-risk that entire information measured system is faced be 0.1 × 0.0591+0.3 × 0.2208+0.5 ×
0.5007+0.7 × 0.1709+0.9 × 0.0485=0.4858.
To sum up, the present invention proposes a kind of opinion rating reliability vector according to the thought of theorem in Euclid space vector projection and is just trading
Shadow decomposition algorithm, which, which can integrate multidigit expertise and can handle expert, provides multiple evaluation knots due to uncertainty
Then the case where fruit, carries out Fuzzy processing to evaluation result by Gauss subordinating degree function again, finally combines Bayesian network
The reasoning algorithm of model solves the risk size that information measured system is faced.The method can enhance assessment result objectivity and
Validity, so that the control and management for subsequent risk provide more reasonable effective foundation.
It is obvious to a person skilled in the art that invention is not limited to the details of the above exemplary embodiments, Er Qie
In the case where without departing substantially from spirit or essential attributes of the invention, the present invention can be realized in other specific forms.Therefore, no matter
From the point of view of which point, the present embodiments are to be considered as illustrative and not restrictive, and the scope of the present invention is by appended power
Benefit requires rather than above description limits, it is intended that all by what is fallen within the meaning and scope of the equivalent elements of the claims
Variation is included within the present invention.Any reference signs in the claims should not be construed as limiting the involved claims.
In addition, it should be understood that although this specification is described in terms of embodiments, but not each embodiment is only wrapped
Containing an independent technical solution, this description of the specification is merely for the sake of clarity, and those skilled in the art should
It considers the specification as a whole, the technical solutions in the various embodiments may also be suitably combined, forms those skilled in the art
The other embodiments being understood that.
Claims (6)
1. a kind of network security risk evaluation method based on fuzzy Bayes, it is characterised in that: this method comprises:
S101, by the evaluation result of the comprehensive multidigit expert of reliability vector rectangular projection decomposition algorithm, calculate each venture influence because
Element is in the reliability of different risk class；
S102, the reliability for being in different risk class to each venture influence factor calculated in S101 quantify, and recycle high
This subordinating degree function carries out Fuzzy processing to the opinion rating quantized value that expert provides, and calculating is under the jurisdiction of different risk class
Degree is weighted summation in conjunction with reliability, calculates the probability value that each venture influence factor is in different risk class；
S103, the probability value that each venture influence factor calculated in S102 is in different risk class is input to by Bayesian network
In the information security risk evaluation index system of network building, the probability value that entire information system is in different risk class is calculated；
S104, basis " gravity model appoach " carry out anti fuzzy method processing, quantify the risk size of entire information system.
2. a kind of network security risk evaluation method based on fuzzy Bayes according to claim 1, it is characterised in that:
The evaluation result by the comprehensive multidigit expert of reliability vector rectangular projection decomposition algorithm in S101, calculates each venture influence factor
Reliability in different risk class, specifically includes:
1) opinion rating result of the K experts to a certain venture influence factor H, is denoted as Vi, wherein i=1,2 ..., K, then by institute
There is evaluation result mutually simultaneously, to be indicated, be denoted as with identification framework ΩBecause N number of in Ω
Element does not include mutually two-by-two, makees following analogy using the thought of theorem in Euclid space:
(1) identification framework Ω is regarded as the N-dimensional vector space comprising N number of reference axis；
(2) the N number of element not included mutually in Ω is regarded as orthogonal reference axis, i.e. x two-by-two1,x2,…,xj,…,xn；
(3) by the opinion rating result V of every expertiRegard a reliability vector v of N-dimensional vector space Ω asi, i=1,2 ...,
K；
If 2), certain opinion rating result ViThere is M element, enables ViCorresponding reliability vector viWith the angle phase of this M reference axis
Deng being 90 degree with reference axis angles other in coordinate system, therefore, as reference axis xjIt is contained in reliability vector viWhen middle, viWith this
The angular separation cosine of one reference axisAs reference axis xjIt is not included in reliability vector viWhen middle, viWith this
The angular separation cosine of one reference axis
3) each reliability vector v, is calculatediMould | | vi| |, i.e., it is after the accuracy rate normalization of every expert as a result, then calculating each
Reliability vector viIn each reference axis xj, wherein j=1,2 ..., n, on rectangular projection decomposition value, i.e.,
4), by each reliability vector viThe cumulative summation of rectangular projection decomposition value in the same reference axis, then renormalization,
Up to this venture influence factor H each opinion rating reliability.
3. a kind of network security risk evaluation method based on fuzzy Bayes according to claim 1, it is characterised in that:
Different risk class locating for each venture influence factor are quantified in S102, recycle Gauss subordinating degree function to expert
The opinion rating quantized value provided carries out Fuzzy processing, calculates the degree for being under the jurisdiction of different risk class, specifically includes:
1) the risk size of a certain venture influence factor, is divided into N number of grade, quantized value is defined on [0,1] section, and risk is got over
Big value is bigger, conversely, quantized value is smaller；
2) N number of Gauss subordinating degree function, is constructed according to N number of risk class
Wherein, the μ represents the center of subordinating degree function, and the selection of central value is determined according to the risk class of division, and
In [0,1] section, it is uniformly distributed as far as possible；The σ indicates the width of subordinating degree function, reflects expert and provides to oneself
Evaluation result uncertainty, σ is bigger, shows that expert is lower to the certainty factor of assessed value；
3) Fuzzy processing, is carried out to the opinion rating quantized value that expert provides by Gauss subordinating degree function, i.e., after quantization
Opinion rating be separately input in N number of subordinating degree function, renormalization can acquire expert to this venture influence factor
Opinion rating is under the jurisdiction of the degree of different risk class.
4. a kind of network security risk evaluation method based on fuzzy Bayes according to claim 1, it is characterised in that:
Combination degree of membership and reliability in S102 are weighted summation, calculate the probability that each venture influence factor is in different risk class
Value, specifically include: the reliability that reliability vector rectangular projection decomposition algorithm is acquired is under the jurisdiction of difference with opinion rating as weight
The degree of risk class is multiplied, then the probability value of same risk class is added and is in different wind to get each venture influence factor
The probability value of dangerous grade.
5. a kind of network security risk evaluation method based on fuzzy Bayes according to claim 1, it is characterised in that:
The probability value that each venture influence factor is in different risk class in S103 is input to the information constructed by Bayesian network
In security risk assessment index system, the probability value that entire information system is in different risk class is calculated, is specifically included:
1), according to information security risk evaluation the relevant technologies and administrative standard, Bayesian network mould is constructed in conjunction with actual conditions
Type；
Wherein, Bayesian network model is made of model structure and model parameter two parts, and model structure is a directed acyclic
Figure, formed by representing the node of variable and representing causal directed arc between variable, model parameter be then represent variable it
Between relationship conditional probability table CPT；
2), Bayesian network model reasoning process specifically includes: if there is n concealed nodes H={ H in a Bayesian network1,
H2,…,HnAnd m Observable node O={ O1,O2,…,Om, node HiFather node be denoted as F (Hi), wherein i=1,2,
3 ..., n, node OjFather node be denoted as F (Oj), wherein j=1,2,3 ..., m, according to conditional independence assumption and d- partition method
Then, the joint probability distribution of all variables is
It, can be according to the probability of the probabilistic inference concealed nodes of observer nodes in conjunction with Bayesian formula
3), using each venture influence factor as the Observable node of Bayesian network model, the risk conduct of entire information system
Each venture influence factor is in the Bayesian network that different grades of probability value is input to building by the root node of Bayesian network
In model, the probability value that entire information system is in different risk class is calculated.
6. a kind of network security risk evaluation method based on fuzzy Bayes according to claim 1, it is characterised in that:
Anti fuzzy method processing is carried out according to " gravity model appoach ", quantifies the risk size of entire information system, specifically includes: setting each risk etc.
The center-of-gravity value of grade, is multiplied, then be added with corresponding probability value, just obtains the risk that entire information measured system is faced
Value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610805770.2A CN106411854B (en) | 2016-09-06 | 2016-09-06 | A kind of network security risk evaluation method based on fuzzy Bayes |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610805770.2A CN106411854B (en) | 2016-09-06 | 2016-09-06 | A kind of network security risk evaluation method based on fuzzy Bayes |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106411854A CN106411854A (en) | 2017-02-15 |
| CN106411854B true CN106411854B (en) | 2019-01-29 |
Family
ID=57998641
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610805770.2A Expired - Fee Related CN106411854B (en) | 2016-09-06 | 2016-09-06 | A kind of network security risk evaluation method based on fuzzy Bayes |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106411854B (en) |
Families Citing this family (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106960269B (en) * | 2017-02-24 | 2021-03-02 | 浙江鹏信信息科技股份有限公司 | Safety emergency disposal method and system based on analytic hierarchy process |
| CN109934447A (en) * | 2018-10-31 | 2019-06-25 | 国网安徽省电力有限公司宿州供电公司 | A Fuzzy Comprehensive Evaluation Method for the Efficiency of Secondary Equipment in Smart Substations |
| CN109978373A (en) * | 2019-03-27 | 2019-07-05 | 金陵科技学院 | A kind of quantification methods of risk assessment based on qualitative data |
| CN110300094B (en) * | 2019-05-08 | 2021-08-03 | 中国人民解放军战略支援部队航天工程大学 | A method, device, device and storage medium for trustworthiness assessment of data nodes |
| CN110598969B (en) * | 2019-06-25 | 2023-03-31 | 大连海事大学 | Offshore channel emergency risk early warning method |
| CN110401649A (en) * | 2019-07-17 | 2019-11-01 | 湖北央中巨石信息技术有限公司 | Information Security Risk Assessment Methods and system based on Situation Awareness study |
| CN110610293A (en) * | 2019-08-13 | 2019-12-24 | 中国人民解放军国防科技大学 | Marine Environmental Risk Assessment Method Based on Improved Bayesian Network |
| CN110968865B (en) * | 2019-11-27 | 2022-03-11 | 桂林电子科技大学 | A risk assessment method for Android software based on probability ontology |
| CN110995692A (en) * | 2019-11-28 | 2020-04-10 | 江苏电力信息技术有限公司 | Network security intrusion detection method based on factor analysis and subspace collaborative representation |
| CN110738399B (en) * | 2019-12-16 | 2023-10-13 | 中山大学 | An early warning method for judicial trial process deviation based on fuzzy set theory |
| CN111476371B (en) * | 2020-06-24 | 2020-09-18 | 支付宝(杭州)信息技术有限公司 | Method and device for evaluating specific risk faced by server |
| CN111932121A (en) * | 2020-08-11 | 2020-11-13 | 广东电网有限责任公司广州供电局 | Method, device, terminal and storage medium for evaluating high-quality power investment scheme |
| CN112232553B (en) * | 2020-09-27 | 2023-04-07 | 同济大学 | Bayesian network-based high-speed rail train late influence factor diagnosis method |
| CN112446643A (en) * | 2020-12-10 | 2021-03-05 | 国网福建省电力有限公司 | Power transmission and transformation project progress risk assessment method based on risk chain |
| CN113379053A (en) * | 2020-12-17 | 2021-09-10 | 中国人民公安大学 | Emergency response decision-making method and device and electronic equipment |
| CN113284340B (en) * | 2021-05-14 | 2022-11-01 | 交通运输部科学研究院 | Road traffic safety grading system |
| CN117749448B (en) * | 2023-12-08 | 2024-05-17 | 广州市融展信息科技有限公司 | Intelligent early warning method and device for network potential risk |
| CN119172162A (en) * | 2024-10-21 | 2024-12-20 | 成都睿的欧科技有限公司 | A network security intelligent assessment method and system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101765231A (en) * | 2009-12-30 | 2010-06-30 | 北京航空航天大学 | Wireless sensor network trust evaluating method based on fuzzy logic |
| CN101763304A (en) * | 2009-12-31 | 2010-06-30 | 合肥工业大学 | Uncertain software credibility assessment method based on evidence theory |
| CN103218689A (en) * | 2013-04-23 | 2013-07-24 | 湖南工学院 | Analyzing method and analyzing device for operator state assessment reliability |
| CN105939200A (en) * | 2015-07-08 | 2016-09-14 | 北京匡恩网络科技有限责任公司 | Method and system for performing network security risk evaluation by utilizing expert system |
-
2016
- 2016-09-06 CN CN201610805770.2A patent/CN106411854B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101765231A (en) * | 2009-12-30 | 2010-06-30 | 北京航空航天大学 | Wireless sensor network trust evaluating method based on fuzzy logic |
| CN101763304A (en) * | 2009-12-31 | 2010-06-30 | 合肥工业大学 | Uncertain software credibility assessment method based on evidence theory |
| CN103218689A (en) * | 2013-04-23 | 2013-07-24 | 湖南工学院 | Analyzing method and analyzing device for operator state assessment reliability |
| CN105939200A (en) * | 2015-07-08 | 2016-09-14 | 北京匡恩网络科技有限责任公司 | Method and system for performing network security risk evaluation by utilizing expert system |
Non-Patent Citations (2)
| Title |
|---|
| e-Risk Management with Insurance : A framework using Copula aided Bayesian;Aruanabha mukhopadhyay 等;《System Sciences, 2006. HICSS "06. Proceedings of the 39th Annual Hawaii International Conference on》;20060107;全文 * |
| 基于贝叶斯网络的信息安全风险评估方法;付钰、吴晓平、严承华;《武汉大学学报（理学版）》;20061031;第52卷(第5期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106411854A (en) | 2017-02-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106411854B (en) | A kind of network security risk evaluation method based on fuzzy Bayes | |
| Brauers et al. | MULTIMOORA for the EU Member States updated with fuzzy number theory | |
| Shiri et al. | Comparison of genetic programming with neuro-fuzzy systems for predicting short-term water table depth fluctuations | |
| Lee et al. | Robust spatial flood vulnerability assessment for Han River using fuzzy TOPSIS with α-cut level set | |
| Warren-Myers | Sustainability evolution in the Australian property market: Examining valuers’ comprehension, knowledge and value | |
| CN106230773A (en) | Risk evaluating system based on fuzzy matrix analytic hierarchy process (AHP) | |
| Ung et al. | A novel human error probability assessment using fuzzy modeling | |
| Arshadi Khamseh et al. | A new fuzzy TOPSIS‐TODIM hybrid method for green supplier selection using fuzzy time function | |
| Evers et al. | Enhancing flood resilience through collaborative modelling and multi-criteria decision analysis (MCDA) | |
| Oshodi et al. | Comparing univariate techniques for tender price index forecasting: Box-Jenkins and neural network model | |
| Bufalo et al. | Improved tourism demand forecasting with CIR# model: a case study of disrupted data patterns in Italy | |
| Jin et al. | Rental price index forecasts of residential properties using Gaussian process regressions | |
| Zeydan et al. | A new hybrid decision making approach for housing suitability mapping of an urban area | |
| Xu et al. | Copula-based high dimensional cross-market dependence modeling | |
| Kung et al. | A fuzzy MCDM method to select the best company based on financial report analysis | |
| Sevgen et al. | Comparison of machine learning algorithms for mass appraisal of real estate data | |
| Uddin et al. | Development of an efficient water quality model using cutting-edge artificial intelligence techniques | |
| Khalilsanjani et al. | Developing a Model Based on Sustainable Development for Prioritizing Entrepreneurial Challenges Under a Competitive Environment. | |
| Dias et al. | Dealing with uncertainty in decision support systems: recent trends (2000–2011) | |
| García-Alonso et al. | A macro-economic model to forecast remittances based on Monte-Carlo simulation and artificial intelligence | |
| Worku | House price drivers in Dubai: nonlinearity and heterogeneity | |
| Padhi et al. | Strategic revenue management under uncertainty: a case study on real estate projects in India | |
| Rana et al. | Best engineering college selection through fuzzy multi-criteria decision making approach: a case study | |
| Scheinert et al. | Growing collaborations: Forecasting changes in partnership networks using a bottom-up approach | |
| Chi Man Hui et al. | Deciphering real estate investment decisions through fuzzy logic systems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20190129 Termination date: 20200906 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |