Office information issuing method based on big data

Abstract

The invention relates to the technical field of office information publishing, in particular to an office information publishing method based on big data. The method comprises the steps of obtaining office information to be published, performing attribute sensitivity calculation on internal business data to obtain attribute sensitivity of the internal business data, constructing a sensitive feature attribute set for the attribute sensitivity based on a hierarchical clustering algorithm, analyzing association rules among attributes of the sensitive feature attribute set through an Apriori algorithm, and obtaining a sensitive feature attribute set; calculating mutual information amount among the attributes with strong association, and constructing an attribute dependency graph; calculating a security index of the external public data based on the external public data; and carrying out sensitive level grading on the internal service data based on the attribute dependency graph to obtain the sensitive level of the internal service data, and judging whether the office information to be published meets a publishing condition based on the sensitive level and the safety index. According to the invention, the leakage of office sensitive information can be prevented, and the security of office information release is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of office information publishing, and in particular to an office information publishing method based on big data. Background Art

[0002] In the era of information explosion, the speed and scope of information dissemination are constantly expanding. Traditional information release methods, such as paper documents, meeting notices, etc., are not only inefficient, but also unable to meet the requirements of information real-time and accuracy. In addition, with the expansion of enterprise scale and business scope, the amount of information has increased dramatically, and traditional processing methods can no longer meet the needs of modern enterprises.

[0003] The rise of big data technology has provided enterprises with a new means of information processing. By collecting, storing, analyzing and mining massive amounts of data, enterprises can more accurately grasp market trends and customer needs, thereby making more scientific and reasonable decisions. In the field of office information release, big data technology also has broad application prospects.

[0004] In the prior art, when office information is released, sensitive information identification and review are often not performed on the internal business data in the office information to be released, resulting in some important information being easily leaked out. Furthermore, security assessment is not performed on the external public data in the office information to be released, resulting in information security risks in the office information to be released.

[0005] Therefore, there is an urgent need for an office information publishing method based on big data to solve the above problems. Summary of the invention

[0006] The purpose of the present invention is to provide an office information publishing method based on big data: to solve the technical problem that in the existing solutions, when office information is released, sensitive information is often not identified and reviewed for internal business data in the office information to be released, which leads to some important information being easily leaked out, and security assessment is not performed on external public data in the office information to be released, resulting in information security risks in the office information to be released.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] The office information publishing method based on big data includes:

[0009] Acquire office information to be published, wherein the office information to be published includes internal business data and external public data;

[0010] Calculate the attribute sensitivity of internal business data to obtain the attribute sensitivity of internal business data, and construct a sensitive feature attribute set based on the attribute sensitivity based on the hierarchical clustering algorithm. Use the Apriori algorithm to analyze the association rules between the attributes of the sensitive feature attribute set, calculate the mutual information between strongly associated attributes, and construct an attribute dependency graph with the associated attributes as vertices, the association relationships between attributes as edges, and the mutual information between attributes as weights.

[0011] Calculate the security index of external public data based on the external public data;

[0012] Based on the attribute dependency graph, the sensitivity level of internal business data is graded to obtain the sensitivity level of the internal business data, and based on the sensitivity level and security index, it is judged whether the office information to be released meets the release conditions. If so, the office information to be released is released, otherwise, the office information to be released is refused to be released.

[0013] Furthermore, the attribute sensitivity of the internal business data is calculated to obtain the attribute sensitivity of the internal business data, which includes the following process:

[0014] Each piece of internal business data is regarded as a source X. Each source X contains n different discrete information (x1, x2, ...x n ), calculate the attribute information entropy H(X) of the source X based on discrete information. The calculation formula of attribute information entropy is as follows:

[0015]

[0016] Among them, p(x i ) represents the probability that discrete information is sensitive information, 0≤p(x i )≤1, and the sum of the probabilities is 1. The probability that discrete information is sensitive information is determined by the keywords in the discrete information.

[0017] Set the maximum discrete entropy of a sensitive attribute to H max (X), calculate the attribute information entropy H of the source X (X and the maximum discrete entropy are H max The Euclidean distance d of (X) i :

[0018] d i =H max (X)-H(X)

[0019] Calculate attribute sensitivity SN i :

[0020]

[0021] Where 0≤SN i <1, the smaller the sensitivity value, the higher the sensitivity.

[0022] Furthermore, the probability that the discrete information is sensitive information is determined by keywords in the discrete information including:

[0023] Based on natural language processing technology, keywords in discrete information are identified, and the sensitive parameters of each keyword are obtained through a search engine. The sum of the sensitive parameters of each keyword is added together as the probability of sensitive information. The sensitive parameters of each keyword are stored in the database of the office information publishing system and obtained through a search engine search.

[0024] Furthermore, constructing a sensitive feature attribute set based on the attribute sensitivity of the hierarchical clustering algorithm includes the following processes:

[0025] Step 1: Take attribute sensitivity as input sample and regard the input sample as a class;

[0026] Step 2: Find the most appropriate division position based on the similarity between samples, so that the similarity between the divided subclasses is minimized;

[0027] Step 3: Determine whether the end condition of the algorithm is met. If so, go to step 4, otherwise, go to step 2, where the end condition is whether the number of iterations is reached.

[0028] Step 4: After the algorithm is finished, several subclass groups are obtained. The average sensitivity of each group is calculated, and the subclass group with the smallest average sensitivity is recorded as the sensitive feature attribute set.

[0029] Furthermore, the Apriori algorithm is used to analyze the association rules between the attributes of the sensitive feature attribute set, and the mutual information between the strongly associated attributes is calculated. The associated attributes are used as vertices, the association relationships between the attributes are used as edges, and the mutual information between the attributes is used as weights. The attribute dependency graph is constructed, which includes the following processes:

[0030] Based on the definition of support threshold ε and confidence threshold γ, the Apriori algorithm is used to find all frequent item sets that meet the minimum support threshold from the sensitive feature attribute set. If sup(A→B)>γ and con(A→B)>ε, there is a strong association between attribute A and attribute B. sup(A→B) represents the support of the association rule between attribute A and attribute B, and con(A→B) represents the confidence of the association rule between attribute A and attribute B.

[0031] Calculate the mutual information I(A;B) between strongly correlated attributes:

[0032] I(A;B)=H(A)-H(A|B

[0033] Among them, H(A) is the information entropy corresponding to attribute A, and H(A|B) is the conditional entropy of attributes A and B;

[0034] Before constructing the attribute dependency graph, the attribute association relationship is described in the form of an adjacency matrix, with the associated attributes as vertices, the association relationship between attributes as edges, and the mutual information between attributes as weights. The form of the adjacency matrix is:

[0035]

[0036] Among them, μ is the mutual information threshold, v,j is the element of the matrix;

[0037] According to the adjacency matrix Z[v][j], the attribute dependency graph G is generated, where the relationship attributes constitute the vertex set V{V1, V2...V k}, the relationship between attributes constitutes the edge set E{E1, E2...E b}, the weight of the edge in the graph is the mutual information between attributes, which indicates the correlation strength between attributes.

[0038] Further, calculating the security index of the external public data based on the external public data includes:

[0039] Obtain the data source IP addresses in the external public data, compare them with the registered data source IP addresses, and obtain the number R of IP addresses that are not registered data source IP addresses among the data source IP addresses in the external public data;

[0040] Obtain the number of encrypted data in the external public data, and calculate the ratio T of the number of encrypted data to the number of data that should be encrypted;

[0041] Obtaining a data upload rate L of external public data, where the data upload rate G of external public data is a probability that the public data is uploaded to the public area network;

[0042] Substituting the number of IP addresses R, the ratio T and the data upload rate L into the security index association formula, the security index LMS is obtained. The security index association formula is as follows:

[0043]

[0044] Among them, Z is the security verification parameter, the weight factors δ and θ, and M is the constant coefficient, which are set by the office information publishing system.

[0045] Furthermore, based on the attribute dependency graph, the sensitivity level of the internal business data is graded, and the sensitivity levels of the internal business data include:

[0046] Sensitive attributes are identified based on the attribute dependency graph, and the corresponding attribute sensitivity is obtained based on the sensitive attributes, and the internal business data is graded according to the attribute sensitivity: when the attribute sensitivity corresponding to the sensitive attribute is less than the first preset threshold, the internal business data is sensitive level 1; when the attribute sensitivity corresponding to the sensitive attribute is greater than the first preset threshold and less than the second preset threshold, the internal business data is sensitive level 2; when the attribute sensitivity corresponding to the sensitive attribute is greater than the second preset threshold, the internal business data is sensitive level 3, wherein the second preset threshold is greater than the first preset threshold.

[0047] Further, judging whether the office information to be released meets the release conditions based on the sensitivity level and the security index includes:

[0048] When the sensitivity level is level 1 or level 2, and the security index is greater than the preset security index threshold, the office information to be released is not allowed to be released;

[0049] When the sensitivity level is level three, office information to be released is not allowed to be released.

[0050] Compared with the existing solutions, the present invention achieves the following beneficial effects:

[0051] The present invention obtains office information to be released, wherein the office information to be released includes internal business data and external public data; performs attribute sensitivity calculation on the internal business data to obtain the attribute sensitivity of the internal business data, and constructs a sensitive feature attribute set based on the attribute sensitivity, analyzes the association rules between the attributes of the sensitive feature attribute set through the Apriori algorithm, calculates the mutual information between the attributes with strong association, and constructs an attribute dependency graph with the associated attributes as vertices, the association relationship between the attributes as edges, and the mutual information between the attributes as weights; calculates the security index of the external public data based on the external public data; performs sensitivity level classification on the internal business data based on the attribute dependency graph to obtain the sensitivity level of the internal business data, and judges whether the office information to be released meets the release conditions based on the sensitivity level and the security index. The present invention can prevent the leakage of office sensitive information and improve the security of office information release. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0053] Figure 1 It is a flow chart of a method for publishing office information based on big data according to an embodiment of the present invention;

[0054] Figure 2It is a flow chart of another office information publishing method based on big data according to an embodiment of the present invention. DETAILED DESCRIPTION

[0055] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0056] In addition, the described features, structures or characteristics may be combined in one or more example embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the example embodiments of the present disclosure. However, those skilled in the art will appreciate that the technical solutions of the present disclosure may be practiced while omitting one or more of the specific details, or other methods, components, steps, etc. may be adopted. In other cases, well-known structures, methods, implementations or operations are not shown or described in detail to avoid obscuring various aspects of the present disclosure.

[0057] This embodiment provides an office information publishing method based on big data. Figure 1 is a flow chart of a method for publishing office information based on big data according to an embodiment of the present invention. Figure 1 As shown, the method comprises the following steps:

[0058] Step S101: obtaining office information to be published, wherein the office information to be published includes internal business data and external public data;

[0059] Step S102: Calculate the attribute sensitivity of the internal business data to obtain the attribute sensitivity of the internal business data;

[0060] Step S103: constructing a sensitive feature attribute set based on the attribute sensitivity using a hierarchical clustering algorithm;

[0061] Step S104: Analyze the association rules between the attributes of the sensitive feature attribute set by the Apriori algorithm, calculate the mutual information between the strongly associated attributes, and construct an attribute dependency graph with the associated attributes as vertices, the association relationships between the attributes as edges, and the mutual information between the attributes as weights;

[0062] Step S105: Calculating the security index of the external public data based on the external public data;

[0063] Step S106: classifying the sensitivity level of the internal business data based on the attribute dependency graph to obtain the sensitivity level of the internal business data;

[0064] Step S107: judging whether the office information to be released meets the release conditions based on the sensitivity level and the security index, if yes, proceeding to step S108, if no, proceeding to step S109;

[0065] Step S108: publishing the office information to be published;

[0066] Step S109: Refuse to publish the office information to be published.

[0067] In summary, the present invention obtains office information to be released, wherein the office information to be released includes internal business data and external public data; performs attribute sensitivity calculation on the internal business data to obtain the attribute sensitivity of the internal business data, and constructs a sensitive feature attribute set based on the attribute sensitivity based on the hierarchical clustering algorithm, analyzes the association rules between the attributes of the sensitive feature attribute set through the Apriori algorithm, calculates the mutual information between strongly associated attributes, and constructs an attribute dependency graph with the associated attributes as vertices, the association relationship between the attributes as edges, and the mutual information between the attributes as weights; calculates the security index of the external public data based on the external public data; performs sensitivity level classification on the internal business data based on the attribute dependency graph to obtain the sensitivity level of the internal business data, and judges whether the office information to be released meets the release conditions based on the sensitivity level and the security index. The present invention can prevent the leakage of sensitive office information and improve the security of office information release.

[0068] In some embodiments, performing attribute sensitivity calculation on internal business data to obtain the attribute sensitivity of the internal business data includes the following process:

[0069] Each piece of internal business data is regarded as a source X. Each source X contains n different discrete information (x1, x2, ...x n ), calculate the attribute information entropy H(X) of the source X based on discrete information. The calculation formula of attribute information entropy is as follows:

[0070]

[0071] Among them, p(x i ) represents the probability that discrete information is sensitive information, 0≤p(x i )≤1, and the sum of the probabilities is 1. The probability that discrete information is sensitive information is determined by the keywords in the discrete information.

[0072] Set the maximum discrete entropy of a sensitive attribute to H max (X), calculate the attribute information entropy H of the source X (X and the maximum discrete entropy are H max The Euclidean distance d of (X) i :

[0073] di =H max (X)-H(X)

[0074] Calculate attribute sensitivity SN i :

[0075]

[0076] Where 0≤SN i <1, the smaller the sensitivity value, the higher the sensitivity.

[0077] Furthermore, the probability that the discrete information is sensitive information is determined by keywords in the discrete information including:

[0078] Based on natural language processing technology, keywords in discrete information are identified, and the sensitive parameters of each keyword are obtained through a search engine. The sum of the sensitive parameters of each keyword is added together as the probability of sensitive information. The sensitive parameters of each keyword are stored in the database of the office information publishing system and obtained through a search engine search.

[0079] In some embodiments, constructing a sensitive feature attribute set based on attribute sensitivity using a hierarchical clustering algorithm includes the following process:

[0080] Step 1: Take attribute sensitivity as input sample and regard the input sample as a class;

[0081] Step 2: Find the most appropriate division position based on the similarity between samples, so that the similarity between the divided subclasses is minimized;

[0082] Step 3: Determine whether the end condition of the algorithm is met. If so, go to step 4, otherwise, go to step 2, where the end condition is whether the number of iterations is reached.

[0083] Step 4: After the algorithm is finished, several subclass groups are obtained. The average sensitivity of each group is calculated, and the subclass group with the smallest average sensitivity is recorded as the sensitive feature attribute set.

[0084] In some embodiments, the association rules between attributes of the sensitive feature attribute set are analyzed by the Apriori algorithm, the mutual information between strongly associated attributes is calculated, and the associated attributes are used as vertices, the association relationships between attributes are used as edges, and the mutual information between attributes is used as weights. The attribute dependency graph is constructed including the following process:

[0085] Based on the definition of support threshold ε and confidence threshold γ, the Apriori algorithm is used to find all frequent item sets that meet the minimum support threshold from the sensitive feature attribute set. If sup(A→B)>γ and con(A→B)>ε, there is a strong association between attribute A and attribute B. sup(A→B) represents the support of the association rule between attribute A and attribute B, and con(A→B) represents the confidence of the association rule between attribute A and attribute B.

[0086] Calculate the mutual information I(A;B) between strongly correlated attributes:

[0087] I(A;B)=H(A)-H(A|B

[0088] Among them, H(A) is the information entropy corresponding to attribute A, and H(A|B) is the conditional entropy of attributes A and B;

[0089] Before constructing the attribute dependency graph, the attribute association relationship is described in the form of an adjacency matrix, with the associated attributes as vertices, the association relationship between attributes as edges, and the mutual information between attributes as weights. The form of the adjacency matrix is:

[0090]

[0091] Among them, μ is the mutual information threshold, v,j is the element of the matrix;

[0092] According to the adjacency matrix Z[v][j], the attribute dependency graph G is generated, where the relationship attributes constitute the vertex set V{V1, V2...V k}, the relationship between attributes constitutes the edge set E{E1, E2...E b}, the weight of the edge in the graph is the mutual information between attributes, which indicates the correlation strength between attributes.

[0093] In some embodiments, in step S105, Figure 2 is a flow chart of another method for publishing office information based on big data according to an embodiment of the present invention. Figure 2 As shown, calculating the security index of external public data based on the external public data includes the following steps:

[0094] Step S201: Acquire the data source IP addresses in the external public data, compare them with the registered data source IP addresses, and obtain the number R of IP addresses that are not registered data source IP addresses in the data source IP addresses in the external public data;

[0095] Step S202: Obtain the number of encrypted data in the external public data, and calculate the ratio T of the number of encrypted data to the number of data to be encrypted;

[0096] Step S203: obtaining a data upload rate L of external public data, where the data upload rate G of external public data is the probability that the public data is uploaded to the public area network;

[0097] Step S204: Substitute the number of IP addresses R, the ratio T and the data upload rate L into the security index association formula to obtain the security index LMS. The security index association formula is as follows:

[0098]

[0099] Among them, Z is the security verification parameter, the weight factors δ and θ, and M is the constant coefficient, which are set by the office information publishing system.

[0100] In some embodiments, the sensitivity level of the internal business data is graded based on the attribute dependency graph, and the sensitivity level of the internal business data is obtained including:

[0101] Sensitive attributes are identified based on the attribute dependency graph, and the corresponding attribute sensitivity is obtained based on the sensitive attributes, and the internal business data is graded according to the attribute sensitivity: when the attribute sensitivity corresponding to the sensitive attribute is less than the first preset threshold, the internal business data is sensitive level 1; when the attribute sensitivity corresponding to the sensitive attribute is greater than the first preset threshold and less than the second preset threshold, the internal business data is sensitive level 2; when the attribute sensitivity corresponding to the sensitive attribute is greater than the second preset threshold, the internal business data is sensitive level 3, wherein the second preset threshold is greater than the first preset threshold.

[0102] In some embodiments, judging whether the office information to be published meets the publishing conditions based on the sensitivity level and the security index includes:

[0103] When the sensitivity level is level 1 or level 2, and the security index is greater than the preset security index threshold, the office information to be released is not allowed to be released;

[0104] When the sensitivity level is level three, office information to be released is not allowed to be released.

[0105] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0106] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0107] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0108] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only some logical function divisions. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0109] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0110] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. An office information publishing method based on big data, characterized in that: Methods include: Acquire office information to be published, wherein the office information to be published includes internal business data and external public data; Calculate the attribute sensitivity of internal business data to obtain the attribute sensitivity of internal business data, and construct a sensitive feature attribute set based on the attribute sensitivity based on the hierarchical clustering algorithm. Use the Apriori algorithm to analyze the association rules between the attributes of the sensitive feature attribute set, calculate the mutual information between strongly associated attributes, and construct an attribute dependency graph with the associated attributes as vertices, the association relationships between attributes as edges, and the mutual information between attributes as weights. Calculate the security index of external public data based on the external public data; Based on the attribute dependency graph, the sensitivity level of internal business data is graded to obtain the sensitivity level of the internal business data, and based on the sensitivity level and security index, it is judged whether the office information to be released meets the release conditions. If so, the office information to be released is released, otherwise, the office information to be released is refused to be released.

2. The method for publishing office information based on big data according to claim 1, characterized in that: Calculating the attribute sensitivity of internal business data to obtain the attribute sensitivity of internal business data includes the following process: Each piece of internal business data is regarded as a source X. Each source X contains n different discrete information (x1, x2, ...x n ), calculate the attribute information entropy H(X) of the source X based on discrete information. The calculation formula of attribute information entropy is as follows: Among them, p(x i ) represents the probability that discrete information is sensitive information, 0≤p(x i )≤1, and the sum of the probabilities is 1. The probability that discrete information is sensitive information is determined by the keywords in the discrete information. Set the maximum discrete entropy of a sensitive attribute to H max (X), calculate the attribute information entropy H of the source X (X and the maximum discrete entropy are H max The Euclidean distance d of (X) i : d i =H max (X)-H(X) Calculate attribute sensitivity SN i : Where 0≤SN i <1, the smaller the sensitivity value, the higher the sensitivity.

3. The method for publishing office information based on big data according to claim 2, characterized in that: The probability that discrete information is sensitive information is determined by the keywords in the discrete information, including: Based on natural language processing technology, keywords in discrete information are identified, and the sensitive parameters of each keyword are obtained through a search engine. The sum of the sensitive parameters of each keyword is added together as the probability of sensitive information. The sensitive parameters of each keyword are stored in the database of the office information publishing system and obtained through a search engine search.

4. The method for publishing office information based on big data according to claim 1, characterized in that: Constructing a sensitive feature attribute set based on the hierarchical clustering algorithm includes the following processes: Step 1: Take attribute sensitivity as input sample and regard the input sample as a class; Step 2: Find the most appropriate division position based on the similarity between samples, so that the similarity between the divided subclasses is minimized; Step 3: Determine whether the end condition of the algorithm is met. If so, go to step 4, otherwise, go to step 2, where the end condition is whether the number of iterations is reached. Step 4: After the algorithm is finished, several subclass groups are obtained. The average sensitivity of each group is calculated, and the subclass group with the smallest average sensitivity is recorded as the sensitive feature attribute set.

5. The method for publishing office information based on big data according to claim 1, characterized in that: The Apriori algorithm is used to analyze the association rules between the attributes of the sensitive feature attribute set, calculate the mutual information between the strongly associated attributes, and construct the attribute dependency graph with the associated attributes as vertices, the association relationships between the attributes as edges, and the mutual information between the attributes as weights. The process includes the following: Based on the definition of support threshold ε and confidence threshold γ, the Apriori algorithm is used to find all frequent item sets that meet the minimum support threshold from the sensitive feature attribute set. If sup(A→B)>γ and con(A→B)>ε, there is a strong association between attribute A and attribute B. sup(A→B) represents the support of the association rule between attribute A and attribute B, and con(A→B) represents the confidence of the association rule between attribute A and attribute B. Calculate the mutual information I(A;B) between strongly correlated attributes: I(A;B)=H(A)-H(A|B) Among them, H(A) is the information entropy corresponding to attribute A, and H(A|B) is the conditional entropy of attributes A and B; Before constructing the attribute dependency graph, the attribute association relationship is described in the form of an adjacency matrix, with the associated attributes as vertices, the association relationship between attributes as edges, and the mutual information between attributes as weights. The form of the adjacency matrix is: Among them, μ is the mutual information threshold, v,j is the element of the matrix; According to the adjacency matrix Z[v][j], the attribute dependency graph G is generated, where the relationship attributes constitute the vertex set V{V1, V2...V k }, the relationship between attributes constitutes the edge set E{E1, E2...E b }, the weight of the edge in the graph is the mutual information between attributes, which indicates the correlation strength between attributes.

6. The method for publishing office information based on big data according to claim 1, characterized in that: The security index of external public data calculated based on external public data includes: Obtain the data source IP addresses in the external public data, compare them with the registered data source IP addresses, and obtain the number R of IP addresses that are not registered data source IP addresses among the data source IP addresses in the external public data; Obtain the number of encrypted data in the external public data, and calculate the ratio T of the number of encrypted data to the number of data that should be encrypted; Obtaining a data upload rate L of external public data, where the data upload rate G of external public data is a probability that the public data is uploaded to the public area network; Substituting the number of IP addresses R, the ratio T and the data upload rate L into the security index association formula, the security index LMS is obtained. The security index association formula is as follows: Among them, Z is the security verification parameter, the weight factors δ and θ, and M is the constant coefficient, which are set by the office information publishing system.

7. The method for publishing office information based on big data according to claim 1, characterized in that: Based on the attribute dependency graph, the sensitivity level of internal business data is graded, and the sensitivity levels of internal business data include: Sensitive attributes are identified based on the attribute dependency graph, and the corresponding attribute sensitivity is obtained based on the sensitive attributes, and the internal business data is graded according to the attribute sensitivity: when the attribute sensitivity corresponding to the sensitive attribute is less than the first preset threshold, the internal business data is sensitive level 1; when the attribute sensitivity corresponding to the sensitive attribute is greater than the first preset threshold and less than the second preset threshold, the internal business data is sensitive level 2; when the attribute sensitivity corresponding to the sensitive attribute is greater than the second preset threshold, the internal business data is sensitive level 3, wherein the second preset threshold is greater than the first preset threshold.

8. The method for publishing office information based on big data according to claim 1, characterized in that: Judging whether the office information to be released meets the release conditions based on the sensitivity level and security index includes: When the sensitivity level is level 1 or level 2, and the security index is greater than the preset security index threshold, the office information to be released is not allowed to be released; When the sensitivity level is level three, office information to be released is not allowed to be released.

Images