Archive management system and method based on KMeans clustering algorithm, and program product

Abstract

The invention provides an archive management system and method based on a KMeans clustering algorithm, and a program product, and relates to the technical field of archive management. The system comprises an information acquisition module, a feature extraction module, a classification module, a management module and a party construction information display module, and the information acquisition module is used for acquiring party construction information archives; the feature extraction module is used for performing feature extraction on the party construction information file to obtain a feature matrix corresponding to the party construction information file; the classification module is used for classifying the party building information archives through a KMeans clustering strategy based on the feature matrix to obtain an archive classification result; the management module is used for performing classified management on the party building information archives according to the archive classification result; and the party building information display module is used for displaying data corresponding to a data calling instruction in the classified and managed party building information file in response to the data calling instruction of the user. In this way, the problems that current party building work is tedious, human influence factors are large, and errors are prone to occurring are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of archive management, and in particular to an archive management system, method and program product based on a KMeans clustering algorithm. Background Art

[0002] Party building is a long and arduous task, and party building also needs to keep pace with the times according to the development and changes of the situation. Networked and information-based party building management can enhance the enthusiasm of party members to participate in party building activities and improve the scientific nature of party building work. Digital party building management can also make party building work clearer and more intuitive, especially in the party building work of hospitals. In order to better carry out medical care and improve the comprehensive quality of party building members, it is necessary to quantify the evaluation indicators of party building work and daily medical care work.

[0003] At present, the evaluation of the work quality of organization members and the arrangement of daily activities are usually still carried out through manual statistics and coordination. However, in the hospital environment, the large flow of patients and diverse and rigorous work content make the implementation of party building work particularly difficult, with problems such as cumbersome work, large human influence factors, and easy errors. Summary of the invention

[0004] In view of this, the purpose of the embodiments of the present application is to provide an archive management system, method and program product based on the KMeans clustering algorithm, which can improve the current problems of party building work, such as cumbersome work, large human influence factors and easy errors.

[0005] In order to achieve the above technical objectives, the technical solutions adopted in this application are as follows:

[0006] In a first aspect, an embodiment of the present application provides a file management system based on a KMeans clustering algorithm, the system comprising:

[0007] Information acquisition module, used to obtain party building information files corresponding to multiple users;

[0008] A feature extraction module is used to extract features from the party building information file to obtain a feature matrix corresponding to the party building information file;

[0009] A classification module is used to classify the party building information archives based on the feature matrix and through the KMeans clustering strategy to obtain archive classification results;

[0010] A management module, used for classifying and managing the party building information archives according to the archive classification results;

[0011] The party building information display module is used to respond to the user's data call instruction and display the data corresponding to the data call instruction in the classified and managed party building information file.

[0012] In combination with the first aspect, in some optional implementations, the feature extraction module is further used to:

[0013] For any party building information file X, determine the entropy corresponding to the party building information file:

[0014]

[0015] In the formula, p i represents the probability distribution P(X=x i )=p i ,i=1,2,…,n, n represents the number of features contained in the party building information file X;

[0016] According to the entropy corresponding to the party building information file, each feature in the party building information file X is weighted to obtain the feature matrix:

[0017]

[0018] In the formula, a ij Represents the element in row i and column j in the feature matrix, j = 1, 2, ..., n, c i represents the features in the party building information file X, |Q| represents the number of party building information files, Represents feature c i The frequency of occurrence in the party building information file X, df(c i ) indicates that all party building information files contain feature c i The number of

[0019] In conjunction with the first aspect, in some optional implementations, the classification module is further used to:

[0020] Randomly select N files as the center points of the party building information files;

[0021] According to the feature matrix, the distance between each file and the center point is calculated, and the minimum distance between each file and the center point is used as the loss function;

[0022] Dividing the party building information archive into a plurality of archive clusters according to the loss function;

[0023] Adjusting the position of the center point according to the loss function until the loss function meets a preset convergence condition;

[0024] When it is determined that the loss function satisfies a preset convergence condition, the multiple file clusters corresponding to the party building information file are used as the file classification results.

[0025] In conjunction with the first aspect, in some optional implementations, the management module is further configured to:

[0026] Obtain various indicator types, including patient satisfaction, outpatient-drug ratio, inter-clinic settlement rate, inter-clinic appointment ratio, and file review rate;

[0027] According to the archive classification results and based on the preset indicator calculation strategy, data processing is performed on the party building information archives of the category corresponding to the indicator type to obtain the indicator calculation results.

[0028] In conjunction with the first aspect, in some optional implementations, the classification module is further used to:

[0029] Randomly determine a value of N within a preset range as a tuning parameter, and repeatedly classify the party building information archive through the classification module to obtain multiple archive classification results, a first parameter and a second parameter, wherein the first parameter represents the sum of the squares of deviations of each feature of the party building information archive in each classification process, and the second parameter represents the sum of the squares of deviations between each category of the party building information archive in each classification process;

[0030] The calculation formula of the first parameter is as follows:

[0031]

[0032] In the formula, A represents the first parameter, R m represents the mth cluster obtained by classification, x represents the files in the mth cluster, and Z m represents the mean distance of the mth cluster, and dist represents the distance between two objects;

[0033] The calculation formula of the second parameter is as follows:

[0034]

[0035] In the formula, B represents the second parameter, V l represents the size of the lth cluster, and y represents the total mean distance of all clusters;

[0036] averaging the first parameters and the second parameters respectively to obtain a first mean value and a second mean value;

[0037] Determine an influencing factor according to the first mean value, the second mean value, and the tuning parameter:

[0038]

[0039] In the formula, P represents the impact factor, A v represents the first mean, B vrepresents the second mean, n represents the total number of archives;

[0040] All values ​​within the preset range are traversed to determine multiple influencing factors, and the tuning parameter corresponding to the maximum value of the multiple influencing factors is used as the target parameter, that is, the final value of N.

[0041] In combination with the first aspect, in some optional implementations, the system further includes:

[0042] Data export module, used to obtain constraints input by users;

[0043] And, calling the data list corresponding to the constraint condition, and generating a data report corresponding to the data list according to the data list.

[0044] In combination with the first aspect, in some optional implementations, the system further includes a user management module, the user management module including a supplier management unit, a physician management unit and a background management unit;

[0045] The supplier management unit is used to store first relevant information of the supplier, wherein the first relevant information includes merchant information, login information, authority information and a historical supply list;

[0046] The physician management unit is used to store second relevant information of each physician, wherein the second relevant information includes indicator data, physician personal information and personal administrative data;

[0047] The background management unit is used to store the personal information and management authority of the system administrator;

[0048] Furthermore, the background management unit is also used to edit the first related information or the second related information pointed to by the target operation instruction based on the target operation instruction issued by the system administrator, and the target operation instruction represents a computer instruction whose instruction content is within the authority managed by the system administrator.

[0049] In combination with the first aspect, in some optional implementations, the system further includes:

[0050] The administrative management module is used to obtain administrative events initiated by the first user account and send the administrative events to the second user account for approval. The administrative events include personnel appointments and dismissals, holiday arrangements, changes in authority, work handovers, and rewards and punishments.

[0051] In a second aspect, the present application also provides a file management method based on the KMeans clustering algorithm, which is applied to the above-mentioned system, and the method includes:

[0052] Through the information acquisition module, obtain the party building information files corresponding to multiple users;

[0053] Perform feature extraction on the party building information file through a feature extraction module to obtain a feature matrix corresponding to the party building information file;

[0054] Based on the feature matrix, the party building information archives are classified by the classification module through the KMeans clustering strategy to obtain the archive classification results;

[0055] According to the archive classification results, the party building information archive is classified and managed through a management module;

[0056] In response to the user's data call instruction, the party building information display module displays the data corresponding to the data call instruction in the classified party building information archive.

[0057] In a third aspect, an embodiment of the present application further provides a computer program product, including a computer program, which implements the above method when executed by a processor.

[0058] The invention adopting the above technical solution has the following advantages:

[0059] In the technical solution provided in this application, the system includes an information acquisition module, a feature extraction module, a classification module, a management module and a party building information display module, wherein the information acquisition module is used to obtain the party building information files corresponding to multiple users; the feature extraction module is used to extract features from the party building information files to obtain the feature matrix corresponding to the party building information files; the classification module is used to classify the party building information files based on the feature matrix through the KMeans clustering strategy to obtain the file classification results; the management module is used to classify the party building information files according to the file classification results; the party building information display module is used to respond to the user's data call instruction and display the data corresponding to the data call instruction in the party building information files after classification management. In this way, by classifying and managing the files of each user, the efficiency of file management is improved, and by displaying the indicator data, the enthusiasm of users to improve their personal literacy is mobilized, the digital management of party building work is realized, and the problems of cumbersome work, large human influence factors and easy errors in the current party building work are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] The present application may be further described by the non-limiting embodiments given in the accompanying drawings. It should be understood that the following drawings only illustrate certain embodiments of the present application and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, other relevant drawings may be obtained based on these drawings without creative effort.

[0061] Figure 1A block diagram of an archive management system based on the KMeans clustering algorithm provided in an embodiment of the present application.

[0062] Figure 2 A flowchart of the archive management method based on the KMeans clustering algorithm provided in the embodiment of the present application.

[0063] Icons: 100- Archives management system based on KMeans clustering algorithm; 101- Information acquisition module; 102- Feature extraction module; 103- Classification module; 104- Management module; 105- Party building information display module. DETAILED DESCRIPTION

[0064] The present application will be described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that in the drawings or descriptions, similar or identical parts use the same figure numbers, and the implementation methods not shown or described in the drawings are forms known to ordinary technicians in the relevant technical field. In the description of this application, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0065] Please refer to Figure 1 The present application embodiment provides a file management system based on the KMeans clustering algorithm, which may include an information acquisition module, a feature extraction module, a classification module, a management module, and a party building information display module. The functions of each module may be as follows:

[0066] Information acquisition module, used to obtain party building information files corresponding to multiple users;

[0067] A feature extraction module is used to extract features from the party building information file to obtain a feature matrix corresponding to the party building information file;

[0068] A classification module is used to classify the party building information archives based on the feature matrix and through the KMeans clustering strategy to obtain archive classification results;

[0069] A management module, used for classifying and managing the party building information archives according to the archive classification results;

[0070] The party building information display module is used to respond to the user's data call instruction and display the data corresponding to the data call instruction in the classified and managed party building information file.

[0071] In this embodiment, the party building information file may include the user's (organization members in the party building management system. This application takes the hospital party building management system as an example, and most of the organization members are doctors) personal information (such as position, name, education, professional title, major, etc.), indicator data (such as patient satisfaction, outpatient drug ratio, inter-clinic settlement rate, inter-clinic appointment ratio, file review rate, etc.), holiday arrangements, etc.

[0072] In this embodiment, the party building information archive can be information entered in advance by the system administrator based on paper archives, data, etc., and changed as various instructions are executed during the operation of the system to ensure the authenticity and real-time nature of the archive data. The acquisition of the party building information archive can be manual entry by the system administrator through input devices (such as mouse, keyboard, etc.); it can also be information contained in the image and voice recognized by the front-end acquisition device (such as camera, microphone, etc.); it can also be copied from other databases that record archive data by initiating a database import instruction by the system administrator. There is no specific limitation on the method of obtaining the party building information archive here.

[0073] In this embodiment, the feature extraction module may also be used for:

[0074] For any party building information file X, determine the entropy corresponding to the party building information file:

[0075]

[0076] In the formula, p i represents the probability distribution P(X=x i )=p i ,i=1,2,…,n, n represents the number of features contained in the party building information file X;

[0077] According to the entropy corresponding to the party building information file, each feature in the party building information file X is weighted to obtain the feature matrix:

[0078]

[0079] In the formula, a ij Represents the element in row i and column j in the feature matrix, j = 1, 2, ..., n, c i represents the features in the party building information file X, |Q| represents the number of party building information files, Represents feature c i The frequency of occurrence in the party building information file X, df(c i ) indicates that all party building information files contain feature c i The number of

[0080] In this way, the feature matrix is ​​constructed

[0081] In this embodiment, during the construction of the feature matrix by the feature extraction module, the entropy calculated by formula (1) characterizes the uncertainty of each feature in the party building information file, and the uncertainty is used as the influencing factor (i.e., weight) of the corresponding feature. The feature is weighted by formula (2), and each feature is normalized during the weighting process, thereby reducing the influence of the total number of characters in the party building information file on the feature weighting result.

[0082] In this embodiment, the classification module is also used for:

[0083] Randomly select N files as the center points of the party building information files;

[0084] According to the feature matrix, the distance between each file and the center point is calculated, and the minimum distance between each file and the center point is used as the loss function;

[0085] Dividing the party building information archive into a plurality of archive clusters according to the loss function;

[0086] Adjusting the position of the center point according to the loss function until the loss function meets a preset convergence condition;

[0087] When it is determined that the loss function satisfies a preset convergence condition, the multiple file clusters corresponding to the party building information file are used as the file classification results.

[0088] In this embodiment, the party building information archives are classified by the KMeans clustering algorithm (i.e., the KMeans clustering strategy). Specifically, N archives are randomly selected as the center points of multiple different clusters (which can be understood as multiple sets of archives of different types) in the party building information archives, and the minimum distance between each archive and each center point is used as the loss function. Among them, the calculation of the minimum distance can be any distance measurement method, such as Euclidean distance, Mahalanobis distance, cosine distance, etc., and this embodiment takes Euclidean distance as an example. For each archive, the elements in the feature matrix of each archive are used as parameters for Euclidean distance calculation, the Euclidean distance between the archive and each center point is calculated, and the archive is assigned to the cluster described by the nearest center point. After calculating the Euclidean distance of each feature and assigning it, for each cluster, the average value of all archives in the cluster is calculated, and the center point is adjusted based on the average value, and the above steps are repeated until the loss function meets the preset convergence condition. Among them, the preset convergence condition can be flexibly set according to user needs, for example, the center points adjusted multiple times all refer to the same archive, or the deviation of the iterative loss function from the previous loss function does not exceed 0.1%, then the loss function is determined to converge. When the loss function is finally determined to converge, the multiple file clusters obtained are used as the file classification results. Among them, by classifying the party building information files in the same file cluster together, you can intuitively get multiple similar or identical files (for example, multiple files refer to doctors in the same department, the doctors corresponding to multiple files are below a certain threshold on the same indicator, and the doctors corresponding to multiple files are highly recognized by patients (that is, high patient satisfaction), etc.), thereby realizing classification management.

[0089] In this embodiment, the management module is also used for:

[0090] Obtain various indicator types, including patient satisfaction, outpatient-drug ratio, inter-clinic settlement rate, inter-clinic appointment ratio, and file review rate;

[0091] According to the archive classification results and based on the preset indicator calculation strategy, data processing is performed on the party building information archives of the category corresponding to the indicator type to obtain the indicator calculation results.

[0092] For example, for the calculation of the patient satisfaction index of doctors in Department A of a certain hospital, the party building information file of all the doctors in Department A can be called in the file classification results, and the patient satisfaction column data in the file can be called. Then, based on the preset indicator calculation strategy (in this example, for the patient satisfaction index, you can first remove multiple highest and lowest values ​​for each doctor's patient satisfaction, and then take the average of the remaining patient satisfaction scores to get the doctor's patient satisfaction. Then, the user satisfaction of each doctor is grouped together, and the highest and lowest values ​​are removed, and the average is taken to get the comprehensive patient satisfaction of the department), data processing of the patient satisfaction indicator is implemented to obtain the indicator calculation result.

[0093] In this embodiment, the classification module is also used for:

[0094] Randomly determine a value of N within a preset range as a tuning parameter, and repeatedly classify the party building information archive through the classification module to obtain multiple archive classification results, a first parameter and a second parameter, wherein the first parameter represents the sum of the squares of deviations of each feature of the party building information archive in each classification process, and the second parameter represents the sum of the squares of deviations between each category of the party building information archive in each classification process;

[0095] The calculation formula of the first parameter is as follows:

[0096]

[0097] In the formula, A represents the first parameter, R m represents the mth cluster obtained by classification, x represents the files in the mth cluster, and Z m represents the mean distance of the mth cluster, and dist represents the distance between two objects;

[0098] The calculation formula of the second parameter is as follows:

[0099]

[0100] In the formula, B represents the second parameter, V l represents the size of the lth cluster, and y represents the total mean distance of all clusters;

[0101] averaging the first parameters and the second parameters respectively to obtain a first mean value and a second mean value;

[0102] Determine an influencing factor according to the first mean value, the second mean value, and the tuning parameter:

[0103]

[0104] In the formula, P represents the impact factor, Av represents the first mean, B v represents the second mean, n represents the total number of archives;

[0105] All values ​​within the preset range are traversed to determine multiple influencing factors, and the tuning parameter corresponding to the maximum value of the multiple influencing factors is used as the target parameter, that is, the final value of N.

[0106] In this embodiment, the preset range can be flexibly set according to actual conditions, such as 3-10, 3-20, etc.

[0107] For example, the preset range can be set to 3-10, and then the value of N is taken from 3 to 10 in sequence. For each value of N, the classification module is used to repeatedly classify the party building information archives one thousand times (so as to avoid local optimal solutions), and then the impact factor corresponding to each value of N is calculated, and the value of N with the largest impact factor is taken as the final value. In this way, the number of iterations of archive classification is reduced, and the burden of data calculation is reduced.

[0108] In this embodiment, the party building information display module is used to display the data that the user wants to call after classification management. For example, it is necessary to display the department with the highest patient satisfaction in the third quarter of xx for display and commendation. According to the user's data call instruction, the data call instruction may include constraints and data flow. For example, the constraint may be to call only patient satisfaction, the time is the third quarter of xx, and the range is for each department. The data flow may be the No. 1 display screen, the No. 2 display screen, and the No. 3 display screen controlled by the party building information display module, and then the party building management module is calculated based on the constraint condition to obtain the index calculation result (i.e., the patient satisfaction of each department), and then the patient satisfaction of each department is sorted, and the department corresponding to the maximum value is selected as the target data, and finally the target data is displayed on the No. 1 display screen, the No. 2 display screen, and the No. 3 display screen to complete the data display.

[0109] As an optional implementation, the system further includes:

[0110] Data export module, used to obtain constraints input by users;

[0111] And, calling the data list corresponding to the constraint condition, and generating a data report corresponding to the data list according to the data list.

[0112] In this embodiment, the data calling process refers to the data calling process involved in the aforementioned data display, and the only difference here is that after calling the data, the data obtained by calling can be used to generate a report according to the user constraint conditions. It is worth mentioning that here it is not limited to traditional text tables, but can also be any form of representing the called data, such as statistical charts, distribution charts, etc.

[0113] As an optional implementation, the system further includes a user management module, which includes a supplier management unit, a physician management unit and a background management unit;

[0114] The supplier management unit is used to store first relevant information of the supplier, wherein the first relevant information includes merchant information, login information, authority information and a historical supply list;

[0115] The physician management unit is used to store second relevant information of each physician, wherein the second relevant information includes indicator data, physician personal information and personal administrative data;

[0116] The background management unit is used to store the personal information and management authority of the system administrator;

[0117] Furthermore, the background management unit is also used to edit the first related information or the second related information pointed to by the target operation instruction based on the target operation instruction issued by the system administrator, and the target operation instruction represents a computer instruction whose instruction content is within the authority managed by the system administrator.

[0118] In this embodiment, editing of relevant information can be understood as adding, deleting, modifying and querying supplier and physician information recorded in this system, such as modifying supplier contact information, inventory of medicines supplied by suppliers, physician contact information, physician title, etc.

[0119] As an optional implementation, the system further includes:

[0120] The administrative management module is used to obtain administrative events initiated by the first user account and send the administrative events to the second user account for approval. The administrative events include personnel appointments and dismissals, holiday arrangements, changes in authority, work handovers, and rewards and punishments.

[0121] In this embodiment, the party building management system can also realize the administrative management of doctors, such as the appointment and removal of doctors, holiday arrangements, changes in authority, work handovers caused by changes in authority or resignation, and various rewards and punishments for doctors.

[0122] Please refer to Figure 2 The present application also provides a file management method based on the KMeans clustering algorithm, which can be applied to the above-mentioned file management system based on the KMeans clustering algorithm. The file management method based on the KMeans clustering algorithm can include the following steps:

[0123] Step 210, obtaining party building information files corresponding to multiple users through the information acquisition module;

[0124] Step 220, extracting features from the party building information file through a feature extraction module to obtain a feature matrix corresponding to the party building information file;

[0125] Step 230, based on the feature matrix, the party building information archive is classified by a classification module through a KMeans clustering strategy to obtain an archive classification result;

[0126] Step 240, classifying and managing the party building information archives through a management module according to the archive classification results;

[0127] Step 250, in response to the user's data call instruction, the party building information display module displays the data corresponding to the data call instruction in the classified party building information file.

[0128] It should be noted that technical personnel in the relevant field can clearly understand that for the convenience and conciseness of description, the specific execution process of the archive management method based on the KMeans clustering algorithm described above can refer to the corresponding operation process or function of each module in the aforementioned system, and will not be elaborated here.

[0129] The present application also provides a computer program product, including a computer program. When the computer program is executed by a processor, the file management method based on the KMeans clustering algorithm described in the above embodiment can be implemented.

[0130] In summary, the embodiment of the present application provides an archive management system, method and program product based on the KMeans clustering algorithm. In this technical solution, the system includes an information acquisition module, a feature extraction module, a classification module, a management module and a party building information display module, wherein the information acquisition module is used to obtain the party building information archives corresponding to multiple users; the feature extraction module is used to extract features from the party building information archives to obtain the feature matrix corresponding to the party building information archives; the classification module is used to classify the party building information archives based on the feature matrix through the KMeans clustering strategy to obtain the archive classification results; the management module is used to classify the party building information archives according to the archive classification results; the party building information display module is used to respond to the user's data call instruction and display the data corresponding to the data call instruction in the party building information archives after classification management. In this way, by classifying and managing the archives of each user, the efficiency of archive management is improved, and by displaying the indicator data, the enthusiasm of users to improve their personal literacy is mobilized, the digital management of party building work is realized, and the problems of cumbersome work, large human influence factors and easy errors in the current party building work are improved.

[0131] In the embodiments provided in the present application, it should be understood that the disclosed system and method can also be implemented in other ways. The system and method embodiments described above are merely schematic, for example, the flow charts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the systems, methods and computer program products according to the multiple embodiments of the present application. In this regard, each box in the flow chart or block diagram can represent a part of a module, a program segment or a code, and a part of the module, program segment or code includes one or more executable instructions for implementing the specified logical function. It should also be noted that each box in the block diagram and / or the flow chart, and the combination of the boxes in the block diagram and / or the flow chart can be implemented by a dedicated hardware-based system that performs the specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions. In addition, each functional module in each embodiment of the present application can be integrated together to form an independent part, or each module can exist separately, or two or more modules can be integrated to form an independent part.

[0132] The above description is only an embodiment of the present application and is not intended to limit the protection scope of the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A file management system based on KMeans clustering algorithm, characterized in that: The system comprises: Information acquisition module, used to obtain party building information files corresponding to multiple users; A feature extraction module is used to extract features from the party building information file to obtain a feature matrix corresponding to the party building information file; A classification module is used to classify the party building information archives based on the feature matrix and through the KMeans clustering strategy to obtain archive classification results; A management module, used for classifying and managing the party building information archives according to the archive classification results; The party building information display module is used to respond to the user's data call instruction and display the data corresponding to the data call instruction in the classified and managed party building information file.

2. The system according to claim 1, characterized in that The feature extraction module is also used for: For any party building information file X, determine the entropy corresponding to the party building information file: In the formula, p i represents the probability distribution P(X=x i )=p i ,i=1,2,…,n, n represents the number of features contained in the party building information file X; According to the entropy corresponding to the party building information file, each feature in the party building information file X is weighted to obtain the feature matrix: In the formula, a ij Represents the element in row i and column j in the feature matrix, j = 1, 2, ..., n, c i represents the features in the party building information file X, |Q| represents the number of party building information files, Represents feature c i The frequency of occurrence in the party building information file X, df(c i ) indicates that all party building information files contain feature c i The number of 3. The system according to claim 1, characterized in that The classification module is also used to: Randomly select N files as the center points of the party building information files; According to the feature matrix, the distance between each file and the center point is calculated, and the minimum distance between each file and the center point is used as the loss function; Dividing the party building information archive into a plurality of archive clusters according to the loss function; Adjusting the position of the center point according to the loss function until the loss function meets a preset convergence condition; When it is determined that the loss function satisfies a preset convergence condition, the multiple file clusters corresponding to the party building information file are used as the file classification results.

4. The system according to claim 1, characterized in that The management module is also used for: Obtain various indicator types, including patient satisfaction, outpatient-drug ratio, inter-clinic settlement rate, inter-clinic appointment ratio, and file review rate; According to the archive classification results and based on the preset indicator calculation strategy, data processing is performed on the party building information archives of the category corresponding to the indicator type to obtain the indicator calculation results.

5. The system according to claim 3, characterized in that The classification module is also used to: Randomly determine a value of N within a preset range as a tuning parameter, and repeatedly classify the party building information archive through the classification module to obtain multiple archive classification results, a first parameter and a second parameter, wherein the first parameter represents the sum of the squares of deviations of each feature of the party building information archive in each classification process, and the second parameter represents the sum of the squares of deviations between each category of the party building information archive in each classification process; The calculation formula of the first parameter is as follows: In the formula, A represents the first parameter, R m represents the mth cluster obtained by classification, x represents the files in the mth cluster, and Z m represents the mean distance of the mth cluster, and dist represents the distance between two objects; The calculation formula of the second parameter is as follows: In the formula, B represents the second parameter, V l represents the size of the lth cluster, and y represents the total mean distance of all clusters; averaging the first parameters and the second parameters respectively to obtain a first mean value and a second mean value; Determine an influencing factor according to the first mean value, the second mean value, and the tuning parameter: In the formula, P represents the impact factor, A v represents the first mean, B v represents the second mean, n represents the total number of archives; All values ​​within the preset range are traversed to determine multiple influencing factors, and the tuning parameter corresponding to the maximum value of the multiple influencing factors is used as the target parameter, that is, the final value of N.

6. The system according to claim 1, characterized in that The system further comprises: Data export module, used to obtain constraints input by users; And, calling the data list corresponding to the constraint condition, and generating a data report corresponding to the data list according to the data list.

7. The system according to claim 1, characterized in that The system further comprises a user management module, which comprises a supplier management unit, a physician management unit and a background management unit; The supplier management unit is used to store first relevant information of the supplier, wherein the first relevant information includes merchant information, login information, authority information and a historical supply list; The physician management unit is used to store second relevant information of each physician, wherein the second relevant information includes indicator data, physician personal information and personal administrative data; The background management unit is used to store the personal information and management authority of the system administrator; Furthermore, the background management unit is also used to edit the first related information or the second related information pointed to by the target operation instruction based on the target operation instruction issued by the system administrator, and the target operation instruction represents a computer instruction whose instruction content is within the authority managed by the system administrator.

8. The system according to claim 1, characterized in that The system further comprises: The administrative management module is used to obtain administrative events initiated by the first user account and send the administrative events to the second user account for approval. The administrative events include personnel appointments and dismissals, holiday arrangements, changes in authority, work handovers, and rewards and punishments.

9. A file management method based on KMeans clustering algorithm, characterized in that: Applied to the system of any one of claims 1 to 8, the method comprising: Through the information acquisition module, obtain the party building information files corresponding to multiple users; Perform feature extraction on the party building information file through a feature extraction module to obtain a feature matrix corresponding to the party building information file; Based on the feature matrix, the party building information archives are classified by the classification module through the KMeans clustering strategy to obtain the archive classification results; According to the archive classification results, the party building information archive is classified and managed through a management module; In response to the user's data call instruction, the party building information display module displays the data corresponding to the data call instruction in the classified party building information archive.

10. A computer program product, characterized in that A computer program is included which, when executed by a processor, implements the method as claimed in claim 9.

Images