CN104809188B - A kind of data mining analysis method of talent drain in corporations and device - Google Patents
A kind of data mining analysis method of talent drain in corporations and device Download PDFInfo
- Publication number
- CN104809188B CN104809188B CN201510187936.4A CN201510187936A CN104809188B CN 104809188 B CN104809188 B CN 104809188B CN 201510187936 A CN201510187936 A CN 201510187936A CN 104809188 B CN104809188 B CN 104809188B
- Authority
- CN
- China
- Prior art keywords
- employee
- data
- cluster
- analysis
- turnover rate
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of data mining analysis method and device of talent drain in corporations, this device at least comprises: data source modules, to being carried out pre-service, cleaning, integrated with conversion by the employee information data collected by Enterprise Salary system and human resource system, and be saved in data collection server; Employees classification module, uses clustering method to carry out cluster analysis to the data of data source modules, employee is carried out cluster, obtains the cluster result of employee; Staff wastage law-analysing module, according to the result of cluster analysis, utilize the Association Rules Model analyzed towards staff wastage, the classified information of employee is associated with the on-job state with running off, causal analysis is carried out to causing the factor of staff wastage, the present invention can carry out analyses and prediction to Mould Enterprise brain drain, obtains brain drain rule, and the turnover rate simultaneously obtaining each employee also can calculate the reference emolument of employee to high turnover rate employee.
Description
Technical field
The present invention relates to data analysis technique field, the data mining analysis method of particularly a kind of Mould Enterprise brain drain and device.
Background technology
The title that mould has " mother of industry " is one of forming method important in manufacturing industry, visible, the critical role that die industry is shared in modern industry.But mould product has feature complicated and changeable, experienced Tool and Die Technology personnel are played an important role in Making mold, so it is indispensable for analyzing mould talent loss in Mould Enterprise.
In enterprise, new employee can make up the loss of old staff wastage, but needs through long-term training adding up with experience, really could substitute the old employee and mould talent of running off.This process will add recruitment cost, the time cost of personnel training and training cost for company.Compare new employee, and the veteran employee in Mould Enterprise is familiar with Order Type and the production status of enterprise self, better can adapt to changeable production process requirements.Retain staff, reducing brain drain rate is the direction that Mould Enterprise Talent Management strategy must be paid close attention to.
But traditional brain drain is analyzed and is mainly taked questionnaire, and the modes such as individual interview are analyzed, and there are some following shortcomings:
The analytical approach that tradition brain drain is analyzed:
1) survey mode, provides questionnaire and carries out wages, the investigation of going or staying purpose etc.;
2) individual interview, carries out interview for the indivedual employees of intentions to remain that have outstanding especially;
3) data of acquisition is carried out Manual analysis statistics and input statistical analysis software or EXCEL office software to analyze.
The shortcoming of above analytical approach:
(1) analysis rule that interview and questionnaire are formulated is subject to the subjective impact of analyst, cannot objectively respond the fact;
(2) because interview needs to mention that employee then can make employee retain to some extent when concrete statement to the idea that interpersonal relation or working environment etc. relate to privacy and there will be the phenomenon of information errors;
(3) interview and questionnaire can obtain the reason of leaving office, but are all that comparison surface is with extensive;
(4) questionnaire content design and be programmed into statistical study questionnaire, needs to drop into very large manpower and materials and runs counter to the high efficiency principle of enterprise.
The human resource management in Mould Enterprise is made to encounter following five problems in brain drain is analyzed by above analytical approach:
The problem that analytic process runs into:
1) when enterprise staff is more, be difficult to carry out Classification Management to employee;
2) present situation describing the existing employee of enterprise from comprehensive angle is difficult to;
3) be difficult to conclude employee's feature of ex-employee;
4) the unpredictable employee possibility of leaving office in a short time;
5) differing greatly due to each employee's classification, reasonably is then difficult to formulate with reference to emolument.
For above-mentioned five problems, traditional management is difficult to the result drawing science.Therefore, the data analysing method by advanced person is needed.Explore and a set ofly can be widely used in Mould Enterprise, and even the data analysing method of manufacturing industry personnel problem is very important.
Summary of the invention
For overcoming the deficiency that above-mentioned prior art exists, the fundamental purpose of the present invention is the data mining analysis method and the device that provide a kind of talent drain in corporations, it passes through from collected employee information data, and integrated, use the data digging method of cluster, the difference degree according to employee's attribute segments, to enable enterprise according to the result of segmentation, deep understanding is carried out to the employee of each classification, and the constitution state of clear understanding drainage of human resources.
Another object of the present invention is the data mining analysis method and the device that provide a kind of talent drain in corporations, it b) carries out staff wastage correlative factor based on correlation rule and probes into, excavate staff wastage rule, emolument welfare can be improved for enterprise, personnel appoint provides reference.
The another object of the present invention is the data mining analysis method and the device that provide a kind of talent drain in corporations, it is predicted by the turnover rate returned based on Logistic, the possibility that prediction employee left office in following a period of time, draw loss probable value, make enterprise administrator can in the analysis result of gained, know which key staff's more likely can be selected to leave office in objective data results in advance, and in advance corresponding excitation made to this part employee and keep measure.
An object again of the present invention is the data mining analysis method and the device that provide talent drain in corporations, it is estimated by the reference emolument based on decision tree, using employee lower for loss probability as the foundation formulated with reference to emolument, and be employee's revaluation emolument of high loss possibility, reduce the reference emolument of turnover rate as one.
For achieving the above object, the data mining analysis device of a kind of talent drain in corporations of the present invention, is characterized in that, this data mining analysis device at least comprises:
Data source modules, to being carried out pre-service, cleaning, integrated with conversion by the employee information data collected by Enterprise Salary system and human resource system, and is saved in data collection server, as the data source of other functional modules;
Employees classification module, uses clustering method to carry out cluster analysis to the data of described data source modules, employee is carried out cluster, obtains the cluster result of employee
Staff wastage law-analysing module, according to the result of cluster analysis, utilizes the Association Rules Model analyzed towards staff wastage, the classified information of employee being associated with the on-job state with running off, carrying out causal analysis to causing the factor of staff wastage.
Further, described data mining analysis device also comprises:
The rate of personnel outflow prediction module, on the basis of cluster analysis, for each cluster sets up a turnover rate forecast model, uses Logistic to return and predicts the rate of personnel outflow.
With reference to emolument module, on the basis obtaining each the rate of personnel outflow, decision-tree model is used to be that the employee that turnover rate is high proposes reasonably with reference to emolument section.
Further, described data source modules comprises data processor and connected workflow engine, data collection server and switch, described workflow engine connects multiple client, described switch is connected with human resource system's data server and emolument system data services device, described data processor utilizes described switch to be connected with the data server of described Enterprise Salary system and human resource system, the request of described data processing timed sending connection data server, after success connects described data server, described data processor sends the services request of extracted data to described data server, after described data server response, the parameter imported into according to described data processor and instruction, copy data in described data processor, described data processor carries out pre-service to the data collected, and extracts, transforms and load data, and pretreated data export and integrate and be loaded into wherein by the model set up according in advance required input variable and output variable, by in the Data import in model to corresponding analysis module for analysis, the data upload handled well is preserved by described data processor in described data collection server.
Further, described employees classification module comprises the classification bulk-breaking of classification engine and connection thereof, described classification engine is connected described data processor with sort file, described sort file merges according to relevant variable linear in employee information, described data processor is made to screen and to filter older historical data, described classification engine is by the model of data stuffing to cluster analysis, and the employee making attributes similarity higher is sub-divided in same cluster, and exports the result of cluster analysis.
Further, described staff wastage law-analysing module comprises correlation rule engine and connected correlation rule interface, described correlation rule engine is connected with described classification engine, on the basis of employee's classification of the human resources of enterprise being carried out segment according to described employees classification module, the data stuffing that described classification engine is obtained to set up in described correlation rule engine in advance towards staff wastage analyze Association Rules Model in, this model by the classified information of employee with on-job with run off state associate; Setting correlation threshold, defines the maximum frequent item variable of association with this; The rule affecting staff wastage is drawn by the model in correlation rule engine.
Further, described the rate of personnel outflow prediction module comprises turnover rate prediction engine and connected turnover rate prediction interface, described turnover rate prediction engine connects described classification engine, in described employees classification module the employee's classification in the human resources of enterprise carried out on the basis segmented, the data stuffing obtained by described classification engine in the model of turnover rate prediction foundation, and sets up a turnover rate forecast model for each cluster; Setting loss early warning line, marks off the boundary of high loss employee and low-bleed employee with this; The loss probable value of each employee is obtained after moving model.
Further, described comprising with reference to emolument module calculates interface with reference to emolument computing engines and connected reference emolument, described reference emolument computing engines connects described turnover rate prediction engine, the turnover rate of each employee can be predicted according to the rate of personnel outflow prediction module, employee is divided into stable type employee and high turnover rate employee; The stable type employee obtained in turnover rate prediction engine is filled into the described decision-tree model set up with reference to emolument computing engines, explores the pay level of stable type employee and the relation of employee's attribute; Carry out new reference emolument to the employee of high turnover rate to calculate, analyze its rationality; The pay level of stable output type employee and the relation of employee's attribute.
For achieving the above object, the present invention also provides a kind of data mining analysis method of talent drain in corporations, comprises the steps:
Step one, to being carried out pre-service, cleaning, integrated with conversion by the employee information data collected by Enterprise Salary system and human resource system, and is saved in data collection server;
Step 2, uses clustering method to carry out cluster analysis to the data in data collection server, employee is carried out cluster, obtains the cluster result of employee;
Step 3, according to the result according to cluster analysis, utilizes the Association Rules Model analyzed towards staff wastage, the classified information of employee being associated with the on-job state with running off, carrying out causal analysis to causing the factor of staff wastage.
Further, described method also comprises the steps:
On the basis of cluster analysis, for each cluster sets up a turnover rate forecast model, use Logistic to return and the rate of personnel outflow is predicted.
Further, described method also comprises the steps:
On the basis obtaining each staff wastage probability, decision-tree model is used to be that the employee that turnover rate is high calculates reasonably with reference to emolument section.
Compared with prior art, the data mining analysis method of a kind of talent drain in corporations of the present invention and device are according to the technology of data mining, first by using cluster analysis that employee is carried out cluster, the characteristic of employee is summarized and classifies, secondly according to correlation rule, causal analysis is carried out to causing the factor of staff wastage, moreover use Logistic recurrence to predict the rate of personnel outflow, the loss orientation of employee is defined, and then carry out remedial measures in advance, decision tree is finally used to be that the employee that turnover rate is high proposes reasonably with reference to emolument section, the invention enables the datumization analytical approach of brain drain generally can be applicable to Mould Enterprise.
Accompanying drawing explanation
Fig. 1 is the system architecture diagram of the data mining analysis device of a kind of talent drain in corporations of the present invention;
Fig. 2 is the structural representation of data source modules 10 in present pre-ferred embodiments;
Fig. 3 is the structural representation of employees classification module 20 in present pre-ferred embodiments;
Fig. 4 is the structural representation of staff wastage law-analysing module 30 in present pre-ferred embodiments;
Fig. 5 is the structural representation of the rate of personnel outflow prediction module 40 in present pre-ferred embodiments;
Fig. 6 is the structural representation with reference to emolument module 50 in present pre-ferred embodiments;
Fig. 7 is the flow chart of steps of the data mining analysis method of a kind of talent drain in corporations of the present invention;
Fig. 8 is the thin portion process flow diagram of the data mining analysis method of talent drain in corporations in present pre-ferred embodiments.
Embodiment
For better the object, technical solutions and advantages of the present invention being described, below in conjunction with the drawings and specific embodiments, the invention will be further described.
Fig. 1 is the system architecture diagram of the data mining analysis device of a kind of talent drain in corporations of the present invention.As shown in Figure 1, the data mining analysis device of a kind of talent drain in corporations of the present invention, comprises data source modules 10, employees classification module 20, staff wastage law-analysing module 30, the rate of personnel outflow prediction module 40 and reference emolument module 50.
Wherein, data source modules 10 to being carried out pre-service, cleaning, integrated with conversion by the employee information data collected by Enterprise Salary system and human resource system, and is saved in data collection server, as the data source of other functional modules; Employees classification module 20, uses clustering method to carry out cluster analysis to the data of data source modules 10, employee is carried out cluster, obtains the cluster result of employee; Staff wastage law-analysing module 30, according to the result of cluster analysis, utilize the Association Rules Model analyzed towards staff wastage, the classified information of employee being associated with the on-job state with running off, carrying out causal analysis to causing the factor of staff wastage; The rate of personnel outflow prediction module 40, on the basis of cluster analysis, for each cluster sets up a turnover rate forecast model, uses Logistic to return and predicts the rate of personnel outflow; With reference to emolument module 50, on the basis obtaining each the rate of personnel outflow, decision-tree model is used to be that the employee that turnover rate is high proposes reasonably with reference to emolument section.
Fig. 2 is the structural representation of data source modules 10 in present pre-ferred embodiments.As shown in Figure 2, data source modules 10, comprise data processor 101 and connected workflow engine 102, data collection server 103 and switch 104, workflow engine 102 is also connected with multiple client 105, and switch 104 is also connected with human resource system's data server 106 and emolument system data services device 107.The operating procedure of data source modules 10 is: use TCP/IP interface to be coupled together with data processor 101 of the present invention by the data server (106/107) of Enterprise Salary system and human resource system with the switch 104 realizing data interaction; The time interval of setting pressed by data processor 101, the request of timed sending connection data server (106/107); After success connection data server (106/107), data processor (101) sends the services request of extracted data to data server (106/107); After data server (106/107) response, the parameter imported into according to data processor (101) and instruction, copy data in data processor (101); Data processor carries out pre-service to the data collected, the process extracting data, transform and load; Pretreated data export and integrate and be loaded into wherein by the model set up according in advance required input variable and output variable; By in the Data import in model to corresponding analysis module for analysis; The data upload handled well is preserved by data processor (101) in data collection server (108).Visible, the function of data source modules 101 is being saved in and analyzing in data source server, as the data source of other functional modules of the present invention, for other module analysis after desired data process.
Fig. 3 is the structural representation of employees classification module 20 in present pre-ferred embodiments.As shown in Figure 3, employees classification module 20, comprise classification engine 201 and with its link sort file 202, sort interface 203 and visualization engine 204, wherein, the data processor 101 of classification engine 201 connection data source module 10, also be connected with visualization interface 205 and display 206 between visualization engine 204 and sort interface 203, server 108 collected by data processor 101 also connection data, and sort file 202 is also connected with data processor 101.Employees classification module 20 operating procedure is: connect employees classification module by interface, classify to employee; Sort file merges according to relevant variable linear in employee information, makes data processor 101 to screen and to filter older historical data; Classification engine 201 is by the model of data stuffing to cluster analysis, and the employee making attributes similarity higher is sub-divided in same cluster; Export the result of cluster analysis, and can view by different work posts, age, academic employee's cluster result for underlying attribute, understand the situation of human resources fast all sidedly; Visual operation is carried out according to the cluster result that above step draws, by visualization engine 204, relevant data are shown according to the employee information etc. of whether leaving office and be correlated with, the classification situation of relevant employee's segmentation can be viewed by link visualization interface.
Fig. 4 is the structural representation of staff wastage law-analysing module 30 in present pre-ferred embodiments.As shown in Figure 4, staff wastage law-analysing module 30 comprises correlation rule engine 3 01 and is connected correlation rule interface 302 and visualization engine 303 with it, wherein, correlation rule engine 3 01 is connected with classification engine 201, also be connected with visualization interface 304 and display 305 between correlation rule interface 302 and visualization engine 303, classification engine 201 is connection data processor 101 also, and server 108 collected by data processor 101 also connection data.The operating procedure of staff wastage law-analysing module is: on the basis of employee's classification of the human resources of enterprise to be carried out segmenting according to employees classification module 20, connect staff wastage law-analysing module by interface, employee is carried out to the analysis of loss law; The data stuffing that classification engine 201 is obtained to set up in correlation rule engine in advance towards staff wastage analyze Association Rules Model in, this model by the classified information of employee with on-job with run off state associate; Setting correlation threshold, defines the maximum frequent item variable of association with this; The important rule affecting staff wastage is drawn by the model in correlation rule engine; User can view relevant staff wastage rule by link visualization interface, and makes relevant behave.
Fig. 5 is the structural representation of the rate of personnel outflow prediction module 40 in present pre-ferred embodiments.As shown in Figure 5, the rate of personnel outflow prediction module 40, comprise turnover rate prediction engine 401 and be connected visualization engine 402 with it and turnover rate predicts interface 403, wherein turnover rate prediction engine 401 link sort engine 201, visualization interface 404 and display 405 is also connected with between turnover rate prediction interface 403 and visualization engine 402, classification engine 201 is connection data processor 101 also, and server 101 collected by data processor 101 also connection data.The operating procedure of the rate of personnel outflow prediction module 40 is: the employee's classification in the human resources of enterprise to be carried out on the basis segmented in employees classification module 20, connect the rate of personnel outflow prediction module 40 by interface, forecast analysis is carried out to the turnover rate of each employee; The data stuffing obtained by classification engine 201 in the model of turnover rate prediction foundation, and sets up a turnover rate forecast model for each cluster; Setting loss early warning line, marks off the boundary of high loss employee and low-bleed employee with this; The loss probable value of each employee will be obtained after moving model; User can view relevant staff wastage rule by link visualization interface, and employee is divided into stable type employee and high turnover rate employee.
Fig. 6 is the structural representation with reference to emolument module 50 in present pre-ferred embodiments.As shown in Figure 6, comprise with reference to emolument computing engines 501 and connected visualization engine 502 with reference to emolument module 50 and calculate interface 503 with reference to emolument, wherein connect turnover rate prediction engine 401 with reference to emolument computing engines 501, calculate between interface 503 and visualization engine 502 with reference to emolument and be also connected with visualization interface 504 and display 505, turnover rate prediction engine is by classification engine 201 connection data processor 101, and server 108 collected by data processor 101 also connection data.Operating procedure with reference to emolument module is: the turnover rate can predicting each employee according to the rate of personnel outflow prediction module, is divided into stable type employee and high turnover rate employee by employee; Connect with reference to emolument computing module by interface; The stable type employee obtained in turnover rate prediction engine is filled into the decision-tree model set up with reference to emolument computing engines, explores the pay level of stable type employee and the relation of employee's attribute; Carry out new reference emolument to the employee of high turnover rate to calculate, analyze its rationality; User can view the pay level of stable type employee and the relation of employee's attribute by link visualization interface, contrasts the association attributes of high turnover rate with this.
Fig. 7 is the flow chart of steps of the data mining analysis method of a kind of talent drain in corporations of the present invention.As shown in Figure 7, the data mining analysis method of a kind of talent drain in corporations of the present invention, comprises the steps:
Step 701, to being carried out pre-service, cleaning, integrated with conversion by the employee information data collected by Enterprise Salary system and human resource system, and is saved in data collection server.Specifically, data processor timing extracted data from Enterprise Salary system and human resource system, according to the data type required for Mould Enterprise human resources situation and data mining analysis, carry out the pre-service of data, cleaning, integrated with conversion, and preserved in the data upload handled well to data collection server.
Step 702, uses clustering method to carry out cluster analysis to the data in data collection server, employee is carried out cluster, obtains the cluster result of employee.Specifically, first employee is presorted, such as employee is presorted as managerial personnel, technician and common employee, then self-defined cluster numbers, perform two step clustering algorithms again to segment employee, solving the difficult problem that there is multiple numerical variable and classified variable when segmenting employee so well, being presented the human resource situation of enterprise by cluster analysis comprehensively.
Step 703, according to the result according to cluster analysis, utilizes the Association Rules Model analyzed towards staff wastage, the classified information of employee being associated with the on-job state with running off, carrying out causal analysis to causing the factor of staff wastage.In this step, relate to a difficult problem for multiple numerical variable and classifying type variable for front step simultaneously, GRI algorithm is adopted to set up the Association Rules Model analyzed towards staff wastage, and set the maximum item number comprised in corresponding preceding paragraph minimum support, minimum regular degree of confidence, preceding paragraph, maximum create-rule number, obtain the important rule affecting staff wastage.
Preferably, after step 702, the data mining analysis method of the present invention also comprises the steps: on the basis of cluster analysis, for each cluster sets up a turnover rate forecast model, uses Logistic to return and predicts the rate of personnel outflow.Specifically, in this step, principal component analysis (PCA) is carried out to the numeric type variable of each cluster, by major component as input variable, respectively the To enterprises the rate of personnel outflow forecast model based on Logistic regretional analysis is set up to each cluster, obtain the turnover rate of each employee by model and set up staff wastage early warning line according to this probability.
Preferably, the data mining analysis method of the present invention also comprises the steps: on the basis obtaining each staff wastage probability, uses decision-tree model to be that the employee that turnover rate is high proposes reasonably with reference to emolument section.Specifically, this step is based on the result of the rate of personnel outflow forecast model, post-class processing analysis is carried out to stability employee, Modling model, excavate the emolument of stable type employee and the rule of their employee's attribute, the rule excavated by post-class processing is applied on high turnover rate employee group, carries out the calculating with reference to emolument to it, encourages the reference frame of way as the salary adjustment of formulating prevention staff wastage.
Fig. 8 is the thin portion process flow diagram of the data mining analysis method of talent drain in corporations in present pre-ferred embodiments.In present pre-ferred embodiments, for Mould Enterprise, the flow process of the data mining analysis method of the Mould Enterprise brain drain of the present invention as shown in Figure 8, data processor timing extracted data from Enterprise Salary system and human resource system, according to the data type required for Mould Enterprise human resources situation and data mining analysis, carry out the pre-service of data, cleaning, integrated with conversion; After data conversion, the actual conditions according to enterprise carry out Variable Selection, obtain managerial personnel, technician and common employee three and presort, perform two step Clustering Model, obtain employee and segment result after self-defined cluster numbers; Result is segmented according to employee, relate to a difficult problem for multiple numerical variable and classifying type variable for previous step simultaneously, use principal component analysis (PCA) disposal route, set the maximum item number comprised in corresponding preceding paragraph minimum support, minimum regular degree of confidence, preceding paragraph, maximum create-rule number, the Association Rules Model analyzed towards staff wastage is set up based on GRI algorithm, obtain the important rule affecting staff wastage, provide reference frame for enterprise can formulate the countermeasure preventing key staff's to run off better; According to the cluster result of employee's segmentation, principal component analysis (PCA) is carried out to the numeric type variable of each cluster, by major component as input variable, respectively based on Logistic regretional analysis, To enterprises the rate of personnel outflow forecast model is set up to each cluster, obtain the probability of each staff wastage by model and set up staff wastage early warning line with this; According to the staff wastage probability calculated, be divided into employee go out high turnover rate employee and stability employee, first with stable type employee for sample carries out post-class processing analysis, this analysis can according to attributes such as educational background, age, the length of service, department, positions, excavate the relation between these input variable and emoluments, just the rule that post-class processing excavates can be applied on high turnover rate employee group after carrying out model adjustment, the calculating with reference to emolument is carried out to it.
The present invention can carry out analyses and prediction to Mould Enterprise brain drain, obtain brain drain rule, the turnover rate simultaneously obtaining each employee also can calculate the reference emolument of employee to high turnover rate employee, making actual contribution, having following three advantages on the whole for keeping the enterprise talent:
(1) efficient data analysis capabilities, the present invention carries out intelligence to the business data of magnanimity and analyzes automatically, especially in Mould Enterprise for a large amount of employee and different information processings, excavate out hiding incidence relation and more hidden attribute more rapidly, improve efficiency and the speed of original data analysis;
(2) comprehensively and accurately coverage, data Real-time Obtaining, performance analysis, the present invention is the use of human resource system, database data in emolument system (some human resource system has comprised emolument system), data cover is to all employees, and information is comprehensive and accurate, and preserves according to unified specification.Analyze data source server and can obtain data in real time for analysis module from system, the present invention can carry out performance analysis to data, facilitates enterprise personnel to tackle the change of drainage of human resources fast;
(3) analysis result is obtained quickly and easily, process visualization.The present invention is simple to operate, and do not need complicated step, ease for use is high, and the employee and the human resource manager's contour level that are applicable to very much human resources use, and directly do not contact with employee so the result obtained is more objective and true in analytic process.
Finally to should be noted that; above embodiment is only in order to illustrate technical scheme of the present invention but not limiting the scope of the invention; although be explained in detail the present invention with reference to preferred embodiment; those of ordinary skill in the art is to be understood that; can modify to technical scheme of the present invention or equivalent replacement, and not depart from essence and the scope of technical solution of the present invention.
Claims (7)
1. a data mining analysis device for talent drain in corporations, is characterized in that, this data mining analysis device at least comprises:
Data source modules, to being carried out pre-service, cleaning, integrated with conversion by the employee information data collected by Enterprise Salary system and human resource system, and is saved in data collection server, as the data source of other functional modules;
Employees classification module, uses clustering method to carry out cluster analysis to the data of described data source modules, employee is carried out cluster, obtains the cluster result of employee;
Staff wastage law-analysing module, according to the result of cluster analysis, utilizes the Association Rules Model analyzed towards staff wastage, the classified information of employee being associated with the on-job state with running off, carrying out causal analysis to causing the factor of staff wastage;
The rate of personnel outflow prediction module, on the basis of cluster analysis, for each cluster sets up a turnover rate forecast model, uses Logistic to return and predicts the rate of personnel outflow;
With reference to emolument module, on the basis obtaining each the rate of personnel outflow, decision-tree model is used to be that the employee that turnover rate is high proposes reasonably with reference to emolument section.
2. the data mining analysis device of talent drain in corporations as claimed in claim 1, it is characterized in that: described data source modules comprises data processor and connected workflow engine, data collection server and switch, described workflow engine connects multiple client, described switch is connected with human resource system's data server and emolument system data services device, described data processor utilizes described switch to be connected with the data server of described Enterprise Salary system and human resource system, the request of described data processing timed sending connection data server, after success connects described data server, described data processor sends the services request of extracted data to described data server, after described data server response, the parameter imported into according to described data processor and instruction, copy data in described data processor, described data processor carries out pre-service to the data collected, and extracts, transforms and load data, and pretreated data export and integrate and be loaded into wherein by the model set up according in advance required input variable and output variable, by in the Data import in model to corresponding analysis module for analysis, the data upload handled well is preserved by described data processor in described data collection server.
3. the data mining analysis device of talent drain in corporations as claimed in claim 2, it is characterized in that: described employees classification module comprises the classification bulk-breaking of classification engine and connection thereof, described classification engine is connected described data processor with sort file, described sort file merges according to relevant variable linear in employee information, described data processor is made to screen and to filter older historical data, described classification engine is by the model of data stuffing to cluster analysis, the employee making attributes similarity higher is sub-divided in same cluster, and export the result of cluster analysis.
4. the data mining analysis device of talent drain in corporations as claimed in claim 3, it is characterized in that: described staff wastage law-analysing module comprises correlation rule engine and connected correlation rule interface, described correlation rule engine is connected with described classification engine, on the basis of employee's classification of the human resources of enterprise being carried out segment according to described employees classification module, the data stuffing that described classification engine is obtained to set up in described correlation rule engine in advance towards staff wastage analyze Association Rules Model in, the classified information of employee associates with the on-job state with running off by this model, setting correlation threshold, defines the maximum frequent item variable of association with this, the rule affecting staff wastage is drawn by the model in correlation rule engine.
5. the data mining analysis device of talent drain in corporations as claimed in claim 3, it is characterized in that: described the rate of personnel outflow prediction module comprises turnover rate prediction engine and connected turnover rate prediction interface, described turnover rate prediction engine connects described classification engine, in described employees classification module the employee's classification in the human resources of enterprise carried out on the basis segmented, the data stuffing obtained by described classification engine in the model of turnover rate prediction foundation, and sets up a turnover rate forecast model for each cluster; Setting loss early warning line, marks off the boundary of high loss employee and low-bleed employee with this; The loss probable value of each employee is obtained after moving model.
6. the data mining analysis device of talent drain in corporations as claimed in claim 5, it is characterized in that: described comprising with reference to emolument module calculates interface with reference to emolument computing engines and connected reference emolument, described reference emolument computing engines connects described turnover rate prediction engine, the turnover rate of each employee can be predicted according to the rate of personnel outflow prediction module, employee is divided into stable type employee and high turnover rate employee; The stable type employee obtained in turnover rate prediction engine is filled into the described decision-tree model set up with reference to emolument computing engines, explores the pay level of stable type employee and the relation of employee's attribute; Carry out new reference emolument to the employee of high turnover rate to calculate, analyze its rationality; The pay level of stable output type employee and the relation of employee's attribute.
7. a data mining analysis method for talent drain in corporations, comprises the steps:
Step one, to being carried out pre-service, cleaning, integrated with conversion by the employee information data collected by Enterprise Salary system and human resource system, and is saved in data collection server;
Step 2, uses clustering method to carry out cluster analysis to the data in data collection server, employee is carried out cluster, obtains the cluster result of employee; On the basis of cluster analysis, for each cluster sets up a turnover rate forecast model, use Logistic to return and the rate of personnel outflow is predicted;
Step 3, according to the result according to cluster analysis, utilizes the Association Rules Model analyzed towards staff wastage, the classified information of employee being associated with the on-job state with running off, carrying out causal analysis to causing the factor of staff wastage;
Step 4, on the basis obtaining each staff wastage probability, uses decision-tree model to be that the employee that turnover rate is high calculates reasonably with reference to emolument section.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510187936.4A CN104809188B (en) | 2015-04-20 | 2015-04-20 | A kind of data mining analysis method of talent drain in corporations and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510187936.4A CN104809188B (en) | 2015-04-20 | 2015-04-20 | A kind of data mining analysis method of talent drain in corporations and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104809188A CN104809188A (en) | 2015-07-29 |
| CN104809188B true CN104809188B (en) | 2016-03-23 |
Family
ID=53694010
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510187936.4A Expired - Fee Related CN104809188B (en) | 2015-04-20 | 2015-04-20 | A kind of data mining analysis method of talent drain in corporations and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104809188B (en) |
Families Citing this family (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105160464A (en) * | 2015-08-24 | 2015-12-16 | 安徽味唯网络科技有限公司 | Method for predicting employee working efficiency and resignation time |
| CN105760957B (en) * | 2016-02-23 | 2017-05-31 | 国元证券股份有限公司 | A kind of Forecasting Methodology of the soft customer revenue of security |
| CN107203822A (en) * | 2016-03-16 | 2017-09-26 | 上海吉贝克信息技术有限公司 | Method and system based on the Logistic security customers attrition predictions returned |
| CN106951569A (en) * | 2017-04-07 | 2017-07-14 | 广州隆德信息科技有限公司 | A kind of data mining analysis management system |
| CN107330577A (en) * | 2017-05-23 | 2017-11-07 | 合肥智权信息科技有限公司 | A kind of company evaluation system and method based on big data |
| CN107609835B (en) * | 2017-07-28 | 2023-04-18 | 国网辽宁省电力有限公司 | Power grid manpower configuration application system and method |
| CN107609064A (en) * | 2017-08-30 | 2018-01-19 | 成都中建科联网络科技有限公司 | Rival's intelligent analysis method based on data mining |
| CN107391767A (en) * | 2017-09-11 | 2017-11-24 | 国网上海市电力公司 | The excavation evaluation system and its method of electrical employee overall qualities test data |
| CN109146116A (en) * | 2018-06-13 | 2019-01-04 | 浙江大学 | A kind of construction method of ability to work model, its calculation method of parameters, and labour's assessment prediction device based on the model |
| JP6418671B1 (en) * | 2018-06-26 | 2018-11-07 | 志村 哲祥 | HR management support system |
| CN109377058A (en) * | 2018-10-26 | 2019-02-22 | 中电科新型智慧城市研究院有限公司 | The enterprise of logic-based regression model moves outside methods of risk assessment |
| CN111369044B (en) * | 2020-02-27 | 2023-06-06 | 腾讯云计算（北京）有限责任公司 | Method, device and computer readable storage medium for estimating churn |
| CN111429114A (en) * | 2020-04-16 | 2020-07-17 | 上海应用技术大学 | Staff loss early warning pre-control mechanism system |
| CN111639814A (en) * | 2020-06-02 | 2020-09-08 | 贝壳技术有限公司 | Method, apparatus, medium, and electronic device for predicting occurrence probability of fluctuating behavior |
| CN113762621A (en) * | 2021-09-09 | 2021-12-07 | 南京领行科技股份有限公司 | Network taxi appointment driver departure prediction method and system |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040039593A1 (en) * | 2002-06-04 | 2004-02-26 | Ramine Eskandari | Managing customer loss using customer value |
| CN102163205A (en) * | 2010-02-21 | 2011-08-24 | 施章祖 | Automatic excavation system for analogous customer groups |
-
2015
- 2015-04-20 CN CN201510187936.4A patent/CN104809188B/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN104809188A (en) | 2015-07-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104809188B (en) | A kind of data mining analysis method of talent drain in corporations and device | |
| López-Robles et al. | Understanding the intellectual structure and evolution of Competitive Intelligence: A bibliometric analysis from 1984 to 2017 | |
| CN108417033B (en) | Expressway traffic accident analysis and prediction method based on multi-dimensional factors | |
| CN106777703A (en) | A kind of bus passenger real-time analyzer and its construction method | |
| CN107517131A (en) | A kind of analysis and early warning method based on log collection | |
| Zhao et al. | Label-less: A semi-automatic labelling tool for kpi anomalies | |
| CN105069025A (en) | Intelligent aggregation visualization and management and control system for big data | |
| CN107273295B (en) | Software problem report classification method based on text chaos | |
| KR101542235B1 (en) | Method for mining of real-time data | |
| CN104679827A (en) | Big data-based public information association method and mining engine | |
| CN105184886A (en) | Cloud data center intelligence inspection system and cloud data center intelligence inspection method | |
| CN111680125A (en) | Litigation case analysis method, litigation case analysis device, computer device, and storage medium | |
| CN116384889A (en) | Intelligent analysis method for information big data based on natural language processing technology | |
| Zhang | Application of data mining technology in digital library. | |
| CN114091866A (en) | Intelligent optimization energy-saving system based on energy consumption convenient combined analysis | |
| He et al. | Intelligent Fault Analysis With AIOps Technology | |
| CN117453764A (en) | Data mining analysis method | |
| Qiao et al. | Constructing a data warehouse based decision support platform for China tourism industry | |
| CN104298751A (en) | Method and device for obtaining internet virtual identity characteristic information | |
| CN114529383A (en) | Method and system for realizing tax payment tracking and tax loss early warning | |
| CN114139747A (en) | AIOps intelligent operation and maintenance system based on artificial intelligence technology | |
| CN207232961U (en) | A kind of system that can be tested in real time to watermark protocol data | |
| Sengupta | Applications of data mining in library & information centres: an overview | |
| CN111291102A (en) | High-performance scale statistical calculation method for government affair data mining | |
| Retnowardhani et al. | Review Study of Business Intelligence to Support Strategic Decision Making |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20160323 Termination date: 20190420 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |