Digital transformation system architecture analysis method based on industry large model

Abstract

The invention relates to the technical field of enterprise transformation management, and discloses a digital transformation system architecture analysis method based on an industry large model, and the method comprises the following steps: collecting enterprise organization structure information; generating two-dimensional structure data according to the acquired enterprise organization structure information; inputting the two-dimensional structure data into a reform model, and outputting an organization reform scheme for changing the enterprise organization structure by the reform model; the two-dimensional structure data and second one-dimensional structure data obtained through organizational transformation scheme sorting are input into a division model, the division model outputs organizational transformation sub-schemes for changing the organizational structure of the enterprise, and the organizational structure of the enterprise can be changed by gradually implementing the sub-schemes; according to the method, the enterprise organization structure information is collected regularly, the enterprise organization transformation scheme implemented in stages is generated, and compared with a single-stage enterprise organization transformation scheme, the method can reduce negative effects on enterprise operation in the organization transformation process to the maximum extent.

Description

Technical Field

[0001] The present invention relates to the technical field of enterprise change management, and more specifically, to an analysis method for digital transformation system architecture based on an industry big model. Background Art

[0002] In a rapidly changing market environment, companies need to adapt quickly to changes. By collecting massive amounts of heterogeneous data from the company's internal and external environments, using deep learning technology, and continuously trying and optimizing in massive amounts of historical data, they can independently generate a series of feasible organizational change plans. However, the implementation of organizational change plans requires the mobilization of a large amount of resources and the adjustment of existing processes, which may have an impact on the normal operations of the company. Summary of the invention

[0003] The purpose of the present invention is to provide an analysis method for the digital transformation architecture based on an industry big model in order to solve the above problems.

[0004] The present invention provides a method for analyzing an industry large model digital transformation system architecture, comprising the following steps:

[0005] Step 100, collecting enterprise organizational structure information, which is collected once at each time sampling point, including department information, personnel information and business process information;

[0006] Step 200, generating two-dimensional structure data from the collected enterprise organizational structure information, the two-dimensional structure data includes a data matrix and a relationship matrix, one unit of the data matrix represents one-dimensional structure data of an independent object, the independent object includes a department, a person and a business process, and one unit of the data matrix only contains the enterprise organizational structure information of the independent object it represents;

[0007] The one-dimensional structure data includes n one-dimensional data items ordered by time, and the t-th one-dimensional data item represents the enterprise organizational structure information collected at the t-th moment;

[0008] Step 300, input the two-dimensional structure data into the change model, the change model includes a first intermediate layer, a second intermediate layer and a first output layer, wherein the first intermediate layer inputs the one-dimensional structure data, outputs the first intermediate representation data to the second intermediate layer, the second intermediate layer also inputs the relationship matrix, the second intermediate layer outputs the second intermediate representation data to the first output layer, and the first output layer outputs the organizational change plan representing the change of the enterprise organizational structure;

[0009] Step 400, input the two-dimensional structure data into the partition model, the partition model includes a third intermediate layer, a fourth intermediate layer, a fifth intermediate layer, a feature fusion layer and a second output layer, wherein the third intermediate layer inputs the one-dimensional structure data No. 1, outputs the third intermediate representation data to the feature fusion layer, the fourth intermediate layer inputs the one-dimensional structure data No. 2 obtained by sorting the organizational change plan, outputs the fourth intermediate representation data to the feature fusion layer, the feature fusion layer outputs the sixth intermediate representation data to the fifth intermediate layer, the fifth intermediate layer outputs the fifth intermediate state to the second output layer, and the second output layer outputs the organizational change sub-plan representing the change of the enterprise's organizational structure, and the gradual implementation of the sub-plan can change the enterprise's organizational structure.

[0010] Furthermore, the time sampling point refers to the moment at which data is collected at a predetermined time interval, and the default value of the predetermined time interval is 24 hours.

[0011] Further, department information includes the department name, functions, a list of business processes in which the department participates, the superior department, a list of department employees, and the department work log;

[0012] Personnel information includes basic information of employees, departments they belong to, superiors, division of labor and responsibilities, business processes they participate in, and employee work logs;

[0013] Business process information includes the steps of the business process, the departments involved, the personnel involved, the specific operation requirements at each department and personnel node, and the process log.

[0014] Furthermore, the element in the i-th row and j-th column of the relationship matrix represents the association between the i-th unit of the data matrix and the independent object represented by the j-th unit. If there is an association, the value of this element of the relationship matrix is ​​1, otherwise it is 0;

[0015] There is a relationship between departments when: there is a management relationship between departments;

[0016] The existence of a relationship between employees means: there is a management relationship between employees;

[0017] There is an association between a department and an employee if: the employee belongs to the department;

[0018] The relationship between a department and a business process refers to: the department performs the business process;

[0019] An association between an employee and a business process means that the employee executes the business process.

[0020] Furthermore, the formula for the first intermediate layer is as follows:

[0021] ;

[0022] ;

[0023] ;

[0024] ;

[0025] in, , , represents the first, second, and third weight parameters, , , represents the first, second and third bias parameters, represents the dot product, , and Respectively represent the first, second, and third intermediate states of the vth unit of the data matrix in the two-dimensional structure data, where Represents the tth data item of the one-dimensional structure data corresponding to the vth unit of the data matrix in the two-dimensional structure data, and Respectively represent the t-th and t-1-th first intermediate representation data of the v-th unit of the data matrix in the two-dimensional structure data, n≥t≥1, n represents the total number of the first data items of the first-dimensional structure data, t=1 , tanh is the hyperbolic tangent function, Represents the sigmoid function;

[0026] The formula for the second intermediate layer is as follows:

[0027] ;

[0028] ;

[0029] ;

[0030] in = , = , represents the attention coefficient of the vth and uth units, represents the attention coefficient of the vth and sth units, represents the normalized attention coefficients of the vth and uth units, represents the set of cells of the data matrix that are associated with the vth cell of the data matrix, The second intermediate representation data representing the vth cell of the data matrix, represents an exponential function with a natural constant as base, represents the trainable attention vector parameter, CONCAT represents the concatenation operation, represents transpose, represents the first two-dimensional recognition weight parameter, LeakyReLU is the LeakyReLU activation function;

[0031] The formula for the first output layer is as follows:

[0032] ;

[0033] in Represents a change vector. The cth component value of the change vector represents the probability value of the cth change strategy. The change strategy with the largest probability value is selected as the output. The change strategy group contains all executable change strategies. A change strategy represents the department setting, management level, business process, personnel transfer, performance appraisal system, and training and development plan of an enterprise. The second intermediate representation data representing the vth cell of the data matrix, represents the set of cells of all data matrices, is the fourth weight parameter, is the fourth bias parameter.

[0034] Furthermore, the steps of training the transformation model include:

[0035] Step 301, initializing the parameters of the transformation model;

[0036] Step 302: Observe the enterprise organizational structure information at time e , the transformation strategy implemented at the e-moment , Current corporate organizational structure information , implement change strategies Rewards ;

[0037] Step 303, then calculate the change strategy error:

[0038] ;

[0039] in represents the change strategy error at time e, represents the first discount factor, , Representing the transformation model input The maximum probability value in the first output vector output when Representing the transformation model input The first output vector outputted when the The probability value of

[0040] ;

[0041] in, , , are the first, second and third weight coefficients, Indicates rewards for organizational structure optimization. , among which , are the fourth and fifth weight coefficients, It represents the difference in the overlap of departmental functions before and after the change. It represents the difference of the management level simplification index before and after the change, and the calculation formula is:

[0042] ;

[0043] ;

[0044] in, represents the total number of departments within the organization, and Respectively represent and The functional set of each department, Represents the number of elements in the intersection of two departmental function sets, It indicates the average number of functions each department has; represents the total number of management levels within the organization, It represents the total number of employees in the organization;

[0045] Reward business process efficiency and execute change strategies After that, the operating efficiency of the enterprise's business process is improved, and the calculation formula is:

[0046] ;

[0047] in represents the number of business processes evaluated, and Respectively represent the implementation of change strategies Before and After The average completion time of a business process;

[0048] It means human resource optimization reward;

[0049] ,in , are the sixth and seventh weight coefficients, It represents the difference of staff redundancy before and after the change. It represents the difference in human resource utilization before and after the change. The calculation formula is:

[0050] ;

[0051] ;

[0052] Indicates The workload of each department, Indicates The average work efficiency of the staff in each department is Indicates The average working hours of staff in each department, Indicates The number of employees in each department, Indicates Department's The actual working time of each employee, Indicates the theoretical available working time of employees;

[0053] Step 304, update the transformation model, the updated formula is as follows:

[0054] ;

[0055] , represents the step size of deep learning, Indicates delivery update;

[0056] Step 305, iterate steps 302-304 until the transformation model converges or the number of iterations reaches a set value.

[0057] Furthermore, the second one-dimensional structure data includes w second data items sorted by time, and the xth second data item represents the second two-dimensional structure data generated by the organizational change plan collected at the xth moment.

[0058] Furthermore, the formula of the third intermediate layer is as follows:

[0059] ;

[0060] in Represents the t-th third intermediate representation data of the v-th unit of the data matrix in the two-dimensional structure data, Represents the t-1th third intermediate representation data of the vth unit of the data matrix in the two-dimensional structure data, Indicates the t-th second data item of the first-dimensional structure data corresponding to the v-th unit of the data matrix in the two-dimensional structure data, and are the fifth and sixth weight parameters, is the fifth bias parameter, express Activation function;

[0061] The formula for the fourth intermediate layer is as follows:

[0062] ;

[0063] in Indicates A fourth intermediate representation data, Indicates A fourth intermediate representation data, represents the tth No.2 data item of No.2 one-dimensional structure data, and are the seventh and eighth weight parameters, is the sixth bias parameter;

[0064] The formula of the feature fusion layer is as follows:

[0065] ;

[0066] in, The sixth intermediate representation data representing the vth unit of the data matrix in the two-dimensional structure data, Represents the vth unit of the data matrix in the two-dimensional structure data A third intermediate representation data, Indicates A fourth intermediate representation data, represents the concatenation function, The total number of data items No. 2 representing the one-dimensional structure data No. 2, represents the sum weight matrix, represents the summation bias parameter;

[0067] The formula for the fifth intermediate layer is as follows:

[0068] ;

[0069] in The fifth intermediate representation data representing the vth cell of the data matrix, is the set of cells of the data matrix that are associated with the vth cell of the data matrix, , , is a learnable scalar parameter, represents the second two-dimensional recognition weight parameter, MLP represents multi-layer perceptron;

[0070] The formula for the second output layer is as follows:

[0071] ;

[0072] in represents a partition vector. The dth component value of the partition vector represents the probability value of the dth partition strategy. The partition strategy with the largest probability value is selected as the output. The partition strategy group contains all executable partition strategies. A partition strategy represents resource allocation, expected time, enterprise department settings, enterprise management level, enterprise business process, enterprise personnel transfer, enterprise performance appraisal system, and enterprise personnel training and development plan. represents the fifth intermediate representation data of the vth unit of the data matrix, M represents the set of all units of the data matrix, is the ninth weight parameter, is the seventh bias parameter.

[0073] Furthermore, the steps of training the partitioning model include:

[0074] Step 401, initializing the parameters of the partitioning model;

[0075] Step 402: observe the comprehensive information including the enterprise organizational structure information and organizational change plan information at time f , the partitioning strategy executed at time f , the enterprise organizational structure information at time f+1 , execute the partitioning strategy Rewards ;

[0076] Step 403, then calculate the partitioning strategy error:

[0077] ;

[0078] in represents the partitioning strategy error at time f, represents the first discount factor, , Represents the partition model input The maximum probability value in the first output vector output when Represents the partition model input The first output vector outputted when the partition strategy The probability value of

[0079] ;

[0080] in, , , , are the eighth, ninth, tenth and eleventh weight coefficients, represents the total number of sub-schemes, It represents the negative impact of the implementation of the sub-plan on the company's operations. The calculation formula is:

[0081] ;

[0082] and Respectively indicate execution of The income of the enterprise before and after each sub-scheme; Indicates the rationality of sub-scheme resource allocation, and the calculation formula is:

[0083] ;

[0084] and Respectively represent The resources required to implement each sub-programme and the resources actually allocated; Indicates the rationality of the sub-scheme time arrangement, and the calculation formula is:

[0085] ;

[0086] in, and Respectively represent The expected time and actual time consumed for the execution of each sub-plan;

[0087] Step 404, update the partition model, the updated formula is as follows:

[0088] ;

[0089] , represents the step size of deep learning, Indicates delivery update;

[0090] Step 405, iterate steps 402-404 until the partition model converges or the number of iterations reaches a set value.

[0091] The present invention provides a computer storage medium for storing computer-readable instructions, which, when read, can execute the aforementioned method for analyzing the digital transformation architecture based on an industry big model.

[0092] The beneficial effect of the present invention is that the present invention generates an enterprise organizational change plan that is implemented in stages by regularly collecting enterprise organizational structure information. Compared with a single-stage enterprise organizational change plan, this method can minimize the negative impact on enterprise operations during the organizational change process. BRIEF DESCRIPTION OF THE DRAWINGS

[0093] Figure 1 It is a flow chart of a method for analyzing an industry big model digital transformation system architecture according to the present invention. DETAILED DESCRIPTION

[0094] The subject matter described herein will now be discussed with reference to example embodiments. It should be understood that the discussion of these embodiments is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the contents of this specification. Each example may omit, replace or add various processes or components as needed. In addition, the features described relative to some examples may also be combined in other examples.

[0095] At least one embodiment of the present invention discloses a method for analyzing an industry large model digital transformation architecture. Figure 1 As shown, the following steps are included:

[0096] Step 100, collecting enterprise organizational structure information, which is collected once at each time sampling point, including department information, personnel information and business process information;

[0097] Department information includes the department name, functions, a list of business processes the department is involved in, the superior department, a list of department employees, and the department work log;

[0098] Personnel information includes basic information of employees, departments they belong to, superiors, division of labor and responsibilities, business processes they participate in, and employee work logs;

[0099] Business process information includes the steps of the business process, the departments involved, the personnel involved, the specific operation requirements at each department and personnel node, and the process log;

[0100] The time sampling point refers to the moment when data is collected at a predetermined time interval;

[0101] In one embodiment of the present invention, the default value of the predetermined time interval is 24 hours.

[0102] Step 200, generating two-dimensional structure data from the collected enterprise organizational structure information, the two-dimensional structure data includes a data matrix and a relationship matrix, one unit of the data matrix represents one-dimensional structure data of an independent object, the independent object includes a department, a person and a business process, and one unit of the data matrix only contains the enterprise organizational structure information of the independent object it represents;

[0103] The element in the i-th row and j-th column of the relationship matrix represents the association between the i-th unit of the data matrix and the independent object represented by the j-th unit. If there is an association, the value of this element of the relationship matrix is ​​1, otherwise it is 0;

[0104] There is a relationship between departments when: there is a management relationship between departments;

[0105] The existence of a relationship between employees means: there is a management relationship between employees;

[0106] There is an association between a department and an employee if: the employee belongs to the department;

[0107] The relationship between a department and a business process refers to: the department performs the business process;

[0108] There is a relationship between an employee and a business process when: the employee performs the business process;

[0109] The one-dimensional structure data includes n one-dimensional data items ordered by time, and the t-th one-dimensional data item represents the enterprise organizational structure information collected at the t-th moment;

[0110] Step 300, input the two-dimensional structure data into the change model, the change model includes a first intermediate layer, a second intermediate layer and a first output layer, wherein the first intermediate layer inputs the one-dimensional structure data, outputs the first intermediate representation data to the second intermediate layer, the second intermediate layer also inputs the relationship matrix, the second intermediate layer outputs the second intermediate representation data to the first output layer, and the first output layer outputs the organizational change plan for changing the enterprise organizational structure, including department settings, management levels, business processes, personnel transfers, performance appraisal systems, and training and development plans;

[0111] In one embodiment of the present invention, the formula of the first intermediate layer is as follows:

[0112]

[0113]

[0114]

[0115]

[0116] in, , , represents the first, second, and third weight parameters, , , represents the first, second and third bias parameters, represents the dot product, , and Respectively represent the first, second, and third intermediate states of the vth unit of the data matrix in the two-dimensional structure data, where Represents the tth data item of the one-dimensional structure data corresponding to the vth unit of the data matrix in the two-dimensional structure data, and Respectively represent the t-th and t-1-th first intermediate representation data of the v-th unit of the data matrix in the two-dimensional structure data, n≥t≥1, n represents the total number of the first data items of the first-dimensional structure data, t=1 , tanh is the hyperbolic tangent function, Represents the sigmoid function.

[0117] In one embodiment of the present invention, the formula of the second intermediate layer is as follows:

[0118]

[0119]

[0120]

[0121] in = , = , represents the attention coefficient of the vth and uth units, represents the attention coefficient of the vth and sth units, represents the normalized attention coefficients of the vth and uth units, represents the set of cells of the data matrix that are associated with the vth cell of the data matrix, The second intermediate representation data representing the vth cell of the data matrix, represents an exponential function with a natural constant as base, represents the trainable attention vector parameter, CONCAT represents the concatenation operation, represents transpose, represents the first two-dimensional recognition weight parameter, and LeakyReLU is the LeakyReLU activation function.

[0122] In one embodiment of the present invention, the formula of the first output layer is as follows:

[0123]

[0124] in Represents a change vector. The cth component value of the change vector represents the probability value of the cth change strategy. The change strategy with the largest probability value is selected as the output. The change strategy group contains all executable change strategies. A change strategy represents the department setting, management level, business process, personnel transfer, performance appraisal system, and training and development plan of an enterprise. The second intermediate representation data representing the vth cell of the data matrix, represents the set of cells of all data matrices, is the fourth weight parameter, is the fourth bias parameter.

[0125] In one embodiment of the present invention, the step of training the transformation model includes:

[0126] Step 301, initializing the parameters of the transformation model;

[0127] Step 302: Observe the enterprise organizational structure information at time e , the transformation strategy implemented at the e-moment , Current corporate organizational structure information , implement change strategies Rewards ;

[0128] Step 303, then calculate the change strategy error:

[0129]

[0130] in represents the change strategy error at time e, represents the first discount factor, , Representing the transformation model input The maximum probability value in the first output vector output when Representing the transformation model input The first output vector outputted when the The probability value of

[0131]

[0132] in, , , are the first, second and third weight coefficients, Indicates rewards for organizational structure optimization. , among which , are the fourth and fifth weight coefficients, It represents the difference in the overlap of departmental functions before and after the change. It represents the difference of the management level simplification index before and after the change, and the calculation formula is:

[0133]

[0134]

[0135] in, represents the total number of departments within the organization, and Respectively represent and The functional set of each department, Represents the number of elements in the intersection of two departmental function sets, It indicates the average number of functions each department has; represents the total number of management levels within the organization, It represents the total number of employees in the organization;

[0136] Reward business process efficiency and execute change strategies After that, the operating efficiency of the enterprise's business process is improved, and the calculation formula is:

[0137]

[0138] in represents the number of business processes evaluated, and Respectively represent the implementation of change strategies Before and After The average completion time of a business process;

[0139] It means human resource optimization reward;

[0140] ,in , are the sixth and seventh weight coefficients, It represents the difference of staff redundancy before and after the change. It represents the difference in human resource utilization before and after the change. The calculation formula is:

[0141]

[0142]

[0143] Indicates The workload of each department, Indicates The average work efficiency of the staff in each department is Indicates The average working hours of the staff in each department, Indicates The number of employees in each department, Indicates Department's The actual working time of each employee, Indicates the theoretical available working time of employees;

[0144] Step 304, update the transformation model, the updated formula is as follows:

[0145]

[0146] , represents the step size of deep learning, Indicates delivery update;

[0147] Step 305, iterate steps 302-304 until the transformation model converges or the number of iterations reaches a set value. The default value of this value is 500.

[0148] Step 400, input the two-dimensional structure data into a partitioning model, the partitioning model includes a third intermediate layer, a fourth intermediate layer, a fifth intermediate layer, a feature fusion layer and a second output layer, wherein the third intermediate layer inputs the first one-dimensional structure data, outputs the third intermediate representation data to the feature fusion layer, the fourth intermediate layer inputs the second one-dimensional structure data, outputs the fourth intermediate representation data to the feature fusion layer, the feature fusion layer outputs the sixth intermediate representation data to the fifth intermediate layer, the fifth intermediate layer outputs the fifth intermediate state to the second output layer, the second output layer outputs an organizational change sub-scheme representing the change of the enterprise organizational structure, the organizational change sub-scheme includes department settings, management levels, business processes, personnel transfers, performance appraisal systems, and training and development plans, and the gradual implementation of the sub-schemes can change the organizational structure of the enterprise;

[0149] In one embodiment of the present invention, the organizational change plans are sorted to obtain second one-dimensional structural data; the second one-dimensional structural data includes w second data items sorted by time, and the xth second data item represents the second two-dimensional structural data generated by the organizational change plan collected at the xth moment.

[0150] In one embodiment of the present invention, the formula of the third intermediate layer is as follows:

[0151]

[0152] in Represents the t-th third intermediate representation data of the v-th unit of the data matrix in the two-dimensional structure data, Represents the t-1th third intermediate representation data of the vth unit of the data matrix in the two-dimensional structure data, Indicates the t-th second data item of the first-dimensional structure data corresponding to the v-th unit of the data matrix in the two-dimensional structure data, and are the fifth and sixth weight parameters, is the fifth bias parameter, express Activation function.

[0153] In one embodiment of the present invention, the formula of the fourth intermediate layer is as follows:

[0154]

[0155] in Indicates A fourth intermediate representation data, Indicates A fourth intermediate representation data, represents the tth No.2 data item of No.2 one-dimensional structure data, and are the seventh and eighth weight parameters, is the sixth bias parameter.

[0156] In one embodiment of the present invention, the formula of the feature fusion layer is as follows:

[0157]

[0158] in, The sixth intermediate representation data representing the vth unit of the data matrix in the two-dimensional structure data, Represents the vth unit of the data matrix in the two-dimensional structure data A third intermediate representation data, Indicates A fourth intermediate representation data, represents the concatenation function, The total number of data items No. 2 representing the one-dimensional structure data No. 2, represents the sum weight matrix, Represents the summation bias parameter.

[0159] In one embodiment of the present invention, the formula of the fifth intermediate layer is as follows:

[0160]

[0161] in The fifth intermediate representation data representing the vth cell of the data matrix, is the set of cells of the data matrix that are associated with the vth cell of the data matrix, , , is a learnable scalar parameter, represents the second two-dimensional recognition weight parameter, and MLP represents a multi-layer perceptron.

[0162] In one embodiment of the present invention, the formula of the second output layer is as follows:

[0163]

[0164] in represents a partition vector. The dth component value of the partition vector represents the probability value of the dth partition strategy. The partition strategy with the largest probability value is selected as the output. The partition strategy group contains all executable partition strategies. A partition strategy represents resource allocation, expected time, enterprise department settings, enterprise management level, enterprise business process, enterprise personnel transfer, enterprise performance appraisal system, and enterprise personnel training and development plan. represents the fifth intermediate representation data of the vth unit of the data matrix, M represents the set of all units of the data matrix, is the ninth weight parameter, is the seventh bias parameter.

[0165] In one embodiment of the present invention, the step of training the partition model includes:

[0166] Step 401, initializing the parameters of the partitioning model;

[0167] Step 402: observe the comprehensive information including the enterprise organizational structure information and organizational change plan information at time f , the partitioning strategy executed at time f , the enterprise organizational structure information at time f+1 , execute the partitioning strategy Rewards ;

[0168] Step 403, then calculate the partitioning strategy error:

[0169]

[0170] in represents the partitioning strategy error at time f, represents the first discount factor, , Represents the partition model input The maximum probability value in the first output vector output when Represents the partition model input The first output vector outputted when the partition strategy The probability value of

[0171]

[0172] in, , , , are the eighth, ninth, tenth and eleventh weight coefficients, represents the total number of sub-schemes, It represents the negative impact of the implementation of the sub-plan on the company's operations. The calculation formula is:

[0173]

[0174] and Respectively indicate execution of The income of the enterprise before and after each sub-scheme; Indicates the rationality of sub-scheme resource allocation, and the calculation formula is:

[0175]

[0176] and Respectively represent The resources required to implement each sub-programme and the resources actually allocated; Indicates the rationality of the sub-scheme time arrangement, and the calculation formula is:

[0177]

[0178] in, and Respectively represent The expected time and actual time consumed for the execution of each sub-plan;

[0179] Step 404, update the partition model, the updated formula is as follows:

[0180]

[0181] , represents the step size of deep learning, Indicates delivery update;

[0182] Step 405, iterate steps 402-404 until the partition model converges or the number of iterations reaches a set value, the default value of which is 500.

[0183] In at least one embodiment of the present invention, a computer storage medium is provided for storing computer-readable instructions, which, when read, can execute the aforementioned analysis method of digital transformation architecture based on an industry big model.

[0184] The above describes an embodiment of the present invention, but this embodiment is not limited to the above-mentioned specific implementation mode. The above-mentioned specific implementation mode is merely illustrative and not restrictive. Under the guidance of this embodiment, ordinary technicians in this field can also make more forms of equivalent embodiments, all of which are within the protection of this embodiment.

Claims

1. A method for analyzing the digital transformation architecture of an industry large model, characterized in that: include: Step 100, collecting enterprise organizational structure information, which is collected once at each time sampling point, including department information, personnel information and business process information; Step 200, generating two-dimensional structure data from the collected enterprise organizational structure information, the two-dimensional structure data includes a data matrix and a relationship matrix, one unit of the data matrix represents one-dimensional structure data of an independent object, the independent object includes a department, a person and a business process, and one unit of the data matrix only contains the enterprise organizational structure information of the independent object it represents; The one-dimensional structure data includes n one-dimensional data items ordered by time, and the t-th one-dimensional data item represents the enterprise organizational structure information collected at the t-th moment; Step 300, input the two-dimensional structure data into the change model, the change model includes a first intermediate layer, a second intermediate layer and a first output layer, wherein the first intermediate layer inputs the one-dimensional structure data, outputs the first intermediate representation data to the second intermediate layer, the second intermediate layer also inputs the relationship matrix, the second intermediate layer outputs the second intermediate representation data to the first output layer, and the first output layer outputs the organizational change plan representing the change of the enterprise organizational structure; Step 400, input the two-dimensional structure data into the partition model, the partition model includes a third intermediate layer, a fourth intermediate layer, a fifth intermediate layer, a feature fusion layer and a second output layer, wherein the third intermediate layer inputs the one-dimensional structure data No. 1, outputs the third intermediate representation data to the feature fusion layer, the fourth intermediate layer inputs the one-dimensional structure data No. 2 obtained by sorting the organizational change plan, outputs the fourth intermediate representation data to the feature fusion layer, the feature fusion layer outputs the sixth intermediate representation data to the fifth intermediate layer, the fifth intermediate layer outputs the fifth intermediate state to the second output layer, and the second output layer outputs the organizational change sub-plan representing the change of the enterprise's organizational structure, and the gradual implementation of the sub-plan can change the enterprise's organizational structure.

2. According to claim 1, a method for analyzing the digital transformation system architecture based on an industry big model is characterized in that: The time sampling point refers to the moment when data is collected at a predetermined time interval. The default value of the predetermined time interval is 24 hours.

3. According to claim 1, a method for analyzing the digital transformation system architecture based on an industry big model is characterized in that: Department information includes the department name, functions, a list of business processes the department is involved in, the superior department, a list of department employees, and the department work log; Personnel information includes basic information of employees, departments they belong to, superiors, division of labor and responsibilities, business processes they participate in, and employee work logs; Business process information includes the steps of the business process, the departments involved, the personnel involved, the specific operation requirements at each department and personnel node, and the process log.

4. According to claim 1, a method for analyzing the digital transformation system architecture based on an industry big model is characterized in that: The element in the i-th row and j-th column of the relationship matrix represents the association between the i-th unit of the data matrix and the independent object represented by the j-th unit. If there is an association, the value of this element of the relationship matrix is ​​1, otherwise it is 0; There is a relationship between departments when: there is a management relationship between departments; The existence of a relationship between employees means: there is a management relationship between employees; There is an association between a department and an employee if: the employee belongs to the department; The relationship between a department and a business process refers to: the department performs the business process; An association between an employee and a business process means that the employee executes the business process.

5. According to claim 1, a method for analyzing the digital transformation system architecture based on an industry big model is characterized in that: The formula for the first intermediate layer is as follows: ； ； ； ； in, , , represents the first, second, and third weight parameters, , , represents the first, second and third bias parameters, represents the dot product, , and Respectively represent the first, second, and third intermediate states of the vth unit of the data matrix in the two-dimensional structure data, where Represents the tth data item of the one-dimensional structure data corresponding to the vth unit of the data matrix in the two-dimensional structure data, and Respectively represent the t-th and t-1-th first intermediate representation data of the v-th unit of the data matrix in the two-dimensional structure data, n≥t≥1, n represents the total number of the first data items of the first-dimensional structure data, t=1 , tanh is the hyperbolic tangent function, Represents the sigmoid function; The formula for the second intermediate layer is as follows: ； ； ； in = , = , represents the attention coefficient of the vth and uth units, represents the attention coefficient of the vth and sth units, represents the normalized attention coefficients of the vth and uth units, represents the set of cells of the data matrix that are associated with the vth cell of the data matrix, The second intermediate representation data representing the vth cell of the data matrix, represents an exponential function with a natural constant as base, represents the trainable attention vector parameter, CONCAT represents the concatenation operation, represents transpose, represents the first two-dimensional recognition weight parameter, LeakyReLU is the LeakyReLU activation function; The formula for the first output layer is as follows: ； in Represents a change vector. The cth component value of the change vector represents the probability value of the cth change strategy. The change strategy with the largest probability value is selected as the output. The change strategy group contains all executable change strategies. A change strategy represents the department setting, management level, business process, personnel transfer, performance appraisal system, and training and development plan of an enterprise. The second intermediate representation data representing the vth cell of the data matrix, represents the set of cells of all data matrices, is the fourth weight parameter, is the fourth bias parameter.

6. The method for analyzing the digital transformation system architecture based on the industry big model according to claim 1 is characterized in that: The steps of the transformation model training include: Step 301, initializing the parameters of the transformation model; Step 302: Observe the enterprise organizational structure information at time e , the transformation strategy implemented at the e-moment , Current corporate organizational structure information , implement change strategies Rewards ; Step 303, then calculate the change strategy error: ； in represents the change strategy error at time e, represents the first discount factor, , Representing the transformation model input The maximum probability value in the first output vector output when Representing the transformation model input The first output vector outputted when the The probability value of ； in, , , are the first, second and third weight coefficients, Indicates rewards for organizational structure optimization. , among which , are the fourth and fifth weight coefficients, It represents the difference in the overlap of departmental functions before and after the change. It represents the difference of the management level simplification index before and after the change, and the calculation formula is: ； ； in, represents the total number of departments within the organization, and Respectively represent and The functional set of each department, Represents the number of elements in the intersection of two departmental function sets, It indicates the average number of functions each department has; represents the total number of management levels within the organization, It represents the total number of employees in the organization; Reward business process efficiency and execute change strategies After that, the operating efficiency of the enterprise's business process is improved, and the calculation formula is: ； in represents the number of business processes evaluated, and Respectively represent the implementation of change strategies Before and After The average completion time of a business process; It means human resource optimization reward; ,in , are the sixth and seventh weight coefficients, It represents the difference of staff redundancy before and after the change. It represents the difference in human resource utilization before and after the change. The calculation formula is: ； ； Indicates The workload of each department, Indicates The average work efficiency of the staff in each department is Indicates The average working hours of the staff in each department, Indicates The number of employees in each department, Indicates Department's The actual working time of each employee, Indicates the theoretical available working time of employees; Step 304, update the transformation model, the updated formula is as follows: ； , represents the step size of deep learning, Indicates delivery update; Step 305, iterate steps 302-304 until the transformation model converges or the number of iterations reaches a set value.

7. The method for analyzing the digital transformation system architecture based on the industry big model according to claim 1 is characterized in that: The No. 2 one-dimensional structure data includes w No. 2 data items sorted by time, and the xth No. 2 data item represents the No. 2 two-dimensional structure data generated by the organizational change plan collected at the xth moment.

8. The method for analyzing the digital transformation architecture based on the industry big model according to claim 1 is characterized in that: The formula for the third intermediate layer is as follows: ； in Represents the t-th third intermediate representation data of the v-th unit of the data matrix in the two-dimensional structure data, Represents the t-1th third intermediate representation data of the vth unit of the data matrix in the two-dimensional structure data, Indicates the t-th second data item of the first-dimensional structure data corresponding to the v-th unit of the data matrix in the two-dimensional structure data, and are the fifth and sixth weight parameters, is the fifth bias parameter, express Activation function; The formula for the fourth intermediate layer is as follows: ； in Indicates A fourth intermediate representation data, Indicates A fourth intermediate representation data, represents the tth No.2 data item of No.2 one-dimensional structure data, and are the seventh and eighth weight parameters, is the sixth bias parameter; The formula of the feature fusion layer is as follows: ； in, The sixth intermediate representation data representing the vth unit of the data matrix in the two-dimensional structure data, Represents the vth unit of the data matrix in the two-dimensional structure data A third intermediate representation data, Indicates A fourth intermediate representation data, represents the concatenation function, The total number of data items No. 2 representing the one-dimensional structure data No. 2, represents the sum weight matrix, represents the summation bias parameter; The formula for the fifth intermediate layer is as follows: ； in The fifth intermediate representation data representing the vth cell of the data matrix, is the set of cells of the data matrix that are associated with the vth cell of the data matrix, , , is a learnable scalar parameter, represents the second two-dimensional recognition weight parameter, MLP represents multi-layer perceptron; The formula for the second output layer is as follows: ； in represents a partition vector. The dth component value of the partition vector represents the probability value of the dth partition strategy. The partition strategy with the largest probability value is selected as the output. The partition strategy group contains all executable partition strategies. A partition strategy represents resource allocation, expected time, enterprise department settings, enterprise management level, enterprise business process, enterprise personnel transfer, enterprise performance appraisal system, and enterprise personnel training and development plan. represents the fifth intermediate representation data of the vth unit of the data matrix, M represents the set of all units of the data matrix, is the ninth weight parameter, is the seventh bias parameter.

9. The method for analyzing the digital transformation system architecture based on the industry big model according to claim 1 is characterized in that: The steps for training the partitioning model include: Step 401, initializing the parameters of the partitioning model; Step 402: observe the comprehensive information including the enterprise organizational structure information and organizational change plan information at time f , the partitioning strategy executed at time f , the enterprise organizational structure information at time f+1 , execute the partitioning strategy Rewards ; Step 403, then calculate the partitioning strategy error: ； in represents the partitioning strategy error at time f, represents the first discount factor, , Represents the partition model input The maximum probability value in the first output vector output when Represents the partition model input The first output vector outputted when the partition strategy The probability value of ； in, , , , are the eighth, ninth, tenth and eleventh weight coefficients, represents the total number of sub-schemes, It represents the negative impact of the implementation of the sub-plan on the company's operations. The calculation formula is: ； and Respectively indicate execution of The income of the enterprise before and after each sub-scheme; Indicates the rationality of sub-scheme resource allocation, and the calculation formula is: ； and Respectively represent The resources required to implement each sub-programme and the resources actually allocated; Indicates the rationality of the sub-scheme time arrangement, and the calculation formula is: ； in, and Respectively represent The expected time and actual time consumed for the execution of each sub-plan; Step 404, update the partition model, the updated formula is as follows: ； , represents the step size of deep learning, Indicates delivery update; Step 405, iterate steps 402-404 until the partition model converges or the number of iterations reaches a set value.

10. A computer storage medium, characterized in that: It is used to store computer-readable instructions, which, when read, can execute a method for analyzing an industry large model digital transformation architecture based on any one of claims 1-9.

Images