CN110378543A - Leaving office Risk Forecast Method, device, computer equipment and storage medium - Google Patents
Leaving office Risk Forecast Method, device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN110378543A CN110378543A CN201810325419.2A CN201810325419A CN110378543A CN 110378543 A CN110378543 A CN 110378543A CN 201810325419 A CN201810325419 A CN 201810325419A CN 110378543 A CN110378543 A CN 110378543A
- Authority
- CN
- China
- Prior art keywords
- employee
- communication behavior
- relational network
- leaving office
- network
- 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.)
- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/10—Office automation; Time management
- G06Q10/105—Human resources
Abstract
The embodiment of the invention discloses a kind of leaving office Risk Forecast Method, device, computer equipment and storage mediums, wherein this method comprises: level information and communication behavior information architecture communication behavior property data base according to employee;Relational network figure is constructed based on the communication behavior property data base, wherein the relational network figure is used to describe the communication behavior between the employee of different stage between the employee of same levels;Feature extraction is carried out to the figure signal in the relational network figure, obtains the tissue signature of the organizational structure;It is trained to obtain leaving office risk forecast model using the method for the tissue signature and machine learning, the leaving office risk forecast model is for carrying out leaving office risk profile.The embodiment of the present invention is not necessarily to human intervention, provides objective leaving office risk analysis prediction result based on artificial intelligence technology, manager is helped to find that organizational change leads to the problem of and is resolved early.
Description
Technical field
The present embodiments relate to field of computer technology more particularly to a kind of leaving office Risk Forecast Methods, device, calculating
Machine equipment and storage medium.
Background technique
Current business circles, the organizational change in enterprise is more frequently and universal, but according to statistics, between the past few decades, nearly 70%
Organization change process all end in failure, therefore, how to carry out organizational change become company manager faced it is important
Problem.
There is no the methods that the intelligentized leaving office risk to employee is predicted in the prior art.Traditional organizational change
Management method is intended to study the cost and generated fastness for how reducing that employee pays organization change, and same
When maximize organization change bring Beneficial Effect, as Lewin propose three step models and Kotter propose eight step plans
Model etc..These are all by seasoned professional by analysis of history result and data statistics, rule of thumb
Out, have certain subjectivity, and can not accomplish to change generate influence carry out quantitative analysis, and by manpower at
This limitation can not analyze comprehensively mass data.In addition, also having hysteresis quality, shortage influences organizational change pre-
Survey ability can not accomplish to provide for a rainy day.
Summary of the invention
The embodiment of the present invention provides a kind of leaving office Risk Forecast Method, device, computer equipment and storage medium, to solve
It in the prior art can not leaving office risk to employee the problem of carrying out intelligent Forecasting.
In a first aspect, the embodiment of the invention provides a kind of leaving office Risk Forecast Methods, this method comprises:
Level information and communication behavior information architecture communication behavior property data base according to employee;
Relational network figure is constructed based on the communication behavior property data base, wherein the relational network figure is for describing
Communication behavior between the employee of different stage between the employee of same levels;
Feature extraction is carried out to the figure signal in the relational network figure, obtains the tissue signature of the organizational structure;
It is trained to obtain leaving office risk forecast model using the method for the tissue signature and machine learning, the leaving office wind
Dangerous prediction model is for carrying out leaving office risk profile.
Second aspect, the embodiment of the invention also provides a kind of leaving office risk profile device, which includes:
Database sharing module, for the level information and communication behavior information architecture communication behavior characteristic according to employee
According to library;
Relational network figure constructs module, for constructing relational network figure based on the communication behavior property data base, wherein
The relational network figure is used to describe the communication behavior between the employee of different stage between the employee of same levels;
Tissue signature's extraction module obtains described for carrying out feature extraction to the figure signal in the relational network figure
The tissue signature of organizational structure;
Risk profile module obtains leaving office risk for being trained using the method for the tissue signature and machine learning
Prediction model, the leaving office risk forecast model is for carrying out leaving office risk profile.
The third aspect, the embodiment of the invention also provides a kind of computer equipments, comprising:
One or more processors;
Storage device, for storing one or more programs,
When one or more of programs are executed by one or more of processors, so that one or more of processing
Device realizes the leaving office Risk Forecast Method as described in any embodiment of the present invention.
Fourth aspect, the embodiment of the invention also provides a kind of computer readable storage mediums, are stored thereon with computer
Program realizes the leaving office Risk Forecast Method as described in any embodiment of the present invention when the program is executed by processor.
The embodiment of the present invention utilizes the level information and communication behavior information architecture relational network figure of employee, and passes through figure letter
Number feature extraction, obtain the tissue signature of organizational structure, be trained using the method for the tissue signature and machine learning
To leaving office risk forecast model, the leaving office risk of employee in organizational structure is predicted by leaving office risk forecast model, without artificial
Intervene, provides objectively analysis prediction result based on artificial intelligence technology, manager is helped to find what organizational change generated early
Problem is simultaneously resolved.
Detailed description of the invention
Fig. 1 is the flow chart for the leaving office Risk Forecast Method that the embodiment of the present invention one provides;
Fig. 2 is the flow chart of leaving office Risk Forecast Method provided by Embodiment 2 of the present invention;
Fig. 3 is the flow chart for the leaving office Risk Forecast Method that the embodiment of the present invention three provides;
Fig. 4 is the structural schematic diagram for the leaving office risk profile device that the embodiment of the present invention four provides;
Fig. 5 is a kind of structural schematic diagram for computer equipment that the embodiment of the present invention five provides.
Specific embodiment
The present invention is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched
The specific embodiment stated is used only for explaining the present invention rather than limiting the invention.It also should be noted that in order to just
Only the parts related to the present invention are shown in description, attached drawing rather than entire infrastructure.
Embodiment one
Fig. 1 is the flow chart for the leaving office Risk Forecast Method that the embodiment of the present invention one provides, and the present embodiment is applicable to pre-
The case where surveying labor turnover risk in organizational structure, this method can be executed by leaving office risk profile device, which can be with
It is realized, and can be integrated in a kind of computer equipment by the way of software and/or hardware.As shown in Figure 1, this method is specifically wrapped
It includes:
S110, level information and communication behavior information architecture communication behavior property data base according to employee.
Its organizational structure situation of different tissues is different, but can be provided with different ranks in each tissue, each
The corresponding rank of a employee.The communication behavior of employee is related to the work performance of employee and working condition, and therefore, the present invention is real
It applies example to analyze by the communication behavior to employee, so that the leaving office risk of the stability and employee to tissue carries out in advance
It surveys.Wherein, the communication behavior may include organizing public avenues of communication to be carried out by mail, real-time communication software etc.
Communication.
Preferably, the level information and communication behavior information architecture communication behavior property data base according to employee, packet
It includes:
Obtain the communication behavior information between organizational structure information, employee's essential information and employee, wherein employee's base
This information includes tissue and employee's rank belonging to employee;
Screening system is carried out according to the communication behavior information between the organizational structure information, employee's essential information and employee
Meter obtains each employee and is different from time that communication behavior occurs between the employee of rank between the employee of same rank
Number;
The object and number of the rank of each employee and generation communication behavior are expressed as feature vector, and stores and arrives data
In library, communication behavior property data base is obtained.Wherein, the database include relevant database (such as MySQL, Oracle) and
Non-relational database (such as Hbase, MongoDB).
Specifically, since organizational structure information and employee's essential information may have variation, for example existing department meets with
To the employee for cancelling or leaving office suddenly, therefore, it is necessary to which organizational structure information, employee's essential information are cleaned and are screened,
No longer valid department is got rid of, and removes the information for the abnormal employee such as leave office, the communication of employee is counted in remaining information
Behavioural information counts its number of communications with same level employee and different stage employee respectively as unit of each employee.?
In a kind of embodiment, such as the number of communications in nearest three months can be counted.
In addition it is also necessary to explanation, can be according to tissue the characteristics of, get rid of some incoherent logical with leaving office risk
Letter behavior, for example, the communication behavior between the employee of the platforms such as human resources or finance department.
S120, relational network figure is constructed based on the communication behavior property data base, wherein the relational network figure is used for
Communication behavior between the employee of different stage between the employee of same levels is described.
Figure is the figure being made of the line of several given points and connection two o'clock, and this figure is certain commonly used to describe
Certain particular kind of relationship between things represents things with point, has between indicating corresponding two things with the line of connection two o'clock this
Relationship.The building purpose of relational network figure is the data structure special using one kind, the communication behavior feature that will be had built up
Initial data in database is combined together, for by description different stage employee between and same levels employee it
Between communication behavior indicate the feature of certain time period inner tissue, building mode can be set according to the characteristics of tissue
It sets.
In a kind of concrete implementation mode, as a preference, can be arranged as follows:
Node in relational network figure indicates communication behavior type, which includes any two different stage
Employee between communication and each rank in communication between employee, correspondingly, the value of relational network figure interior joint is by this
The number of communications that the communication behavior type that node indicates occurs indicates;
Side in relational network figure indicates the simultaneous probability of different communication behavior type, correspondingly, relational network figure
It is middle while weight by possessing this simultaneously while corresponding two nodes represented by the headcount of communication behavior type indicate.
Specifically, relational network chart can be shown as to G=<V, E>, wherein each node v ∈ V.Node viAnd vjBetween
Side e ∈ E, weight wijCalculation it is as follows:
Wherein, k and θ is parameter preset, and sim (i, j) indicates the co-occurrence between two kinds of communication behavior types, i.e., occurs simultaneously
The probability of both communication behavior types, representation method include but is not limited to the employee by possessing both communication types simultaneously
Quantity indicates.For example, it is assumed that in some tissue, there are two types of the position A and B of rank, and between A and A, between A and B
Communication is very frequent, then a kind of building mode of relational network figure is, which includes three nodes A-B, A-A and B-B, and A-
The nodal value of A and A-B can be relatively larger.
S130, feature extraction is carried out to the figure signal in the relational network figure, the tissue for obtaining the organizational structure is special
Sign.
Due to having co-occurrence between the node in relational network figure, the local relation in neighbor node can be fine
Help extract the characteristic information contained in the upper figure signal of relational network figure (graph signal).Wherein, in relational network figure
Local relation method for digging include but is not limited to Complex Networks Analysis method, social influence power model IC mode and LT mode,
Max-flow-minimal cut algorithm and figure convolution method etc..The feature for extracting figure signal is used as the tissue signature of organizational structure,
For training leaving office risk forecast model.
S140, it is trained to obtain leaving office risk forecast model using the method for the tissue signature and machine learning, it should
Leaving office risk forecast model is for carrying out leaving office risk profile.
It analyzes to obtain tissue spy according to aforesaid operations specifically, can use history communication behavior data interior for a period of time
Sign utilizes the method progress of machine learning using the leaving office risk situation of this tissue as output as input historical stage
Training obtains leaving office risk forecast model, and prediction target includes but is not limited to the separation rate of tissue, leaving office number and leaving office ring
Than/data such as amplification on year-on-year basis.Wherein it is possible to be trained using neural network, including convolutional neural networks CNN or circulation nerve
Network RNN, further, it is also possible to utilize support vector machines, condition random field CRF, logistic regression LR, random forest random
Forest scheduling algorithm carries out model training.
The technical solution of the present embodiment utilizes the level information and communication behavior information architecture relational network figure of employee, and leads to
The feature extraction for crossing figure signal obtains the tissue signature of organizational structure, is carried out using the method for the tissue signature and machine learning
Training obtains leaving office risk forecast model, and the leaving office risk of employee in organizational structure, nothing are predicted by leaving office risk forecast model
Human intervention is needed, objectively analysis prediction result is provided based on artificial intelligence technology, manager is helped to find organizational change early
It leads to the problem of and is resolved.
Embodiment two
Fig. 2 is the flow chart of leaving office Risk Forecast Method provided by Embodiment 2 of the present invention, and the present embodiment is in above-mentioned reality
Further progress optimizes on the basis of applying example.As shown in Fig. 2, this method comprises:
S210, level information and communication behavior information architecture communication behavior property data base according to employee.
S220, relational network figure is constructed based on the communication behavior property data base, wherein the relational network figure is used for
Communication behavior between the employee of different stage between the employee of same levels is described.
S230, the relational network figure is analyzed using figure convolution operation, obtains the feature of figure signal, which is believed
Number tissue signature of the feature as the organizational structure, wherein the figure signal is by each node in the relational network figure
The vector of value composition indicates.
Specifically, the embodiment of the present invention is to be analyzed using figure convolution operation relational network figure, picture scroll product is to be based on
Spectrogram theory carries out convolution operation to relational network figure, since picture scroll product is it can be considered that integrally-built spy in relational network figure
Sign, therefore feature extraction is more comprehensively and accurate.The Laplacian Matrix L of including but not limited to normalization figure can be used to indicate
Relational network figure, and the figure convolution operation is indicated by carrying out Fourier transformation realization to figure signal are as follows:
Wherein, cGIndicate figure signal, UTIndicate the transposition of the eigenvectors matrix of Laplacian Matrix L.
In addition, the requirement of local feature is excavated in order to reduce complexity and meet convolution, it can be further to convolution
Operation is modified, and specific method includes but is not limited to use multinomial filter hθTo replace convolution operator.Such as K dimension part
Multinomial filter, is defined as:
Wherein,Representative polynomial coefficient, with relational network figure interior joint viCentered on filter interior joint vj
Value by formulaIt calculates, if viAnd vjDistance be more than k (parameter preset), then haveBecause picture scroll product is the Fourier transformation carried out to figure signal, with the formula come calculate node
Value, has just obtained revised figure signal, then Fourier transformation is carried out to revised figure signal, to extract figure signal
Feature.
Wherein,It is a k rank Chebyshev polynomials, by formulaIt calculates.Qie Bixue
Husband's multinomial is that can have T with the sequence of the orthogonal polynomial of recursive definitionn+1(x)=2xTn(x)-Tn-1(x), wherein T0(x)=1
And T1(x)=x.It is widely used in approximation theory, while also reducing the time cost of convolution.
In addition, pondization operation can also be carried out after convolution operation, it is possible to reduce characteristic dimension reduces and calculates time complexity
Degree.Specifically, the clustering algorithm of Subgraph can be adopted to operate as the pondization.
S240, it is trained to obtain leaving office risk forecast model using the method for the tissue signature and machine learning, it should
Leaving office risk forecast model is for carrying out leaving office risk profile.
The technical solution of the present embodiment extracts the feature of figure signal in relational network figure using figure convolution operation, due to figure
Convolution is it can be considered that integrally-built feature in relational network figure, therefore feature extraction is more comprehensively and accurate.Using extracting
Tissue signature and the method for machine learning be trained to obtain leaving office risk forecast model, it is pre- by leaving office risk forecast model
The leaving office risk of employee in organizational structure is surveyed, human intervention is not necessarily to, provides objectively analysis prediction knot based on artificial intelligence technology
Fruit helps manager to find that organizational change leads to the problem of and is resolved early.
Embodiment three
Fig. 3 is the flow chart for the leaving office Risk Forecast Method that the embodiment of the present invention three provides, and the present embodiment is in above-mentioned reality
Further progress optimizes on the basis of applying example.As shown in figure 3, this method comprises:
S310, level information and communication behavior information architecture communication behavior property data base according to employee, the communication
The communication behavior feature of different time slice is stored in behavioral characteristic database.
S320, the communication behavior feature construction relational network figure based on different time slice, wherein the network of personal connections
Network figure is used to describe the communication behavior between the employee of different stage between the employee of same levels.
S330, to the figure signal in the relational network figure for the communication behavior feature construction being sliced according to the different time into
Row feature extraction obtains the tissue signature of the organizational structure.
S340, using the different time slice communication behavior feature corresponding tissue signature and machine learning method
It is trained to obtain leaving office risk forecast model, the leaving office risk forecast model is for carrying out leaving office risk profile.
The technical solution of the present embodiment is unlike the embodiments above, introduces timing information in the present embodiment.Tool
Body, after the level information and communication behavior information for obtaining employee, these communication behavior information are sliced first, are obtained
To the communication behavior information of different slices.Then the communication according to method described in any of the above-described embodiment, according to different slices
Behavioural information constructs communication behavior property data base, constructs relational network figure according to the communication behavior property data base, then base
In relational network figure extraction figure signal characteristic and as tissue signature, it is trained using the tissue signature, obtains leaving office wind
Dangerous prediction model.
And due to different time points, the working strength of employee is in the presence of variation, the work shape of employee in organizational structure
State is also not unalterable, therefore, by introducing timing information, is sliced to communication behavior, to extract different slices
The feature of communication behavior is learnt, and not only may learn the feature of the communication behavior of a certain slice, can also learn difference
The relationship and difference of communication behavior between slice, thus obtain more preferably, more accurate learning effect, improve model prediction
Accuracy.
Wherein, the method for slice includes equidistant slice or equivalent slice.Equidistant slice refers to according to certain frequency progress
Cutting, for example, carrying out cutting in the communication behavior data that will acquire according to one month time interval, obtaining every month
Communication behavior data slicer.Equivalent slice, which refers to, carries out cutting according to number of communications, in the different slices obtained after cutting, respectively
It is identical for being sliced corresponding number of communications all., can according to the actual situation when realization, the different dicing method of option and installment.
Communication behavior data are sliced, and by the introducing of timing information to slice by the technical solution of the present embodiment
Data afterwards handle and learn respectively, thus obtain more preferably, more accurate learning effect, improve the accuracy of model prediction,
Preferably manager is helped to find that organizational change leads to the problem of and is resolved early.
Example IV
Fig. 4 is the structural schematic diagram for the leaving office risk profile device that the embodiment of the present invention four provides, and the present embodiment is applicable
In prediction organizational structure the case where labor turnover risk.Leaving office risk profile device provided by the embodiment of the present invention is executable
Leaving office Risk Forecast Method provided by any embodiment of the invention has the corresponding functional module of execution method and beneficial to effect
Fruit.As shown in figure 4, the device includes database sharing module 410, relational network figure building module 420, tissue signature's extraction mould
Block 430 and risk profile module 440, in which:
Database sharing module 410, the level information and communication behavior information architecture communication behavior for foundation employee are special
Levy database;
Relational network figure constructs module 420, for constructing relational network figure based on the communication behavior property data base,
In, the relational network figure is used to describe the communication behavior between the employee of different stage between the employee of same levels;
Tissue signature's extraction module 430 obtains institute for carrying out feature extraction to the figure signal in the relational network figure
State the tissue signature of organizational structure;
Risk profile module 440 is left office for being trained using the method for the tissue signature and machine learning
Risk forecast model, the leaving office risk forecast model is for carrying out leaving office risk profile.
On the basis of upper technical solution, further, database sharing module 410 includes:
Acquiring unit, for obtaining the communication behavior information between organizational structure information, employee's essential information and employee,
In, employee's essential information includes tissue and employee's rank belonging to employee;
Statistic unit is screened, for according to the communication row between the organizational structure information, employee's essential information and employee
Screening statistics is carried out for information, each employee is obtained and is different between the employee of rank between the employee of same rank
The number of communication behavior occurs;
Storage unit, for by the rank of each employee and occur communication behavior object and number be expressed as feature to
Amount, and store into database, obtain communication behavior property data base.
Further, the node in the relational network figure indicates that communication behavior type, the communication behavior type include
Communication in communication and each rank between the employee of any two different stage between employee, correspondingly, the relationship
Number of communications that communication behavior type that the value of network interior joint is indicated by the node occurs indicates;
Side in the relational network figure indicates the simultaneous probability of different communication behavior type, correspondingly, the pass
Be in network while weight by possessing this simultaneously while corresponding two nodes represented by communication behavior type headcount
To indicate.
Further, tissue signature's extraction module 430 is specifically used for:
The relational network figure is analyzed using figure convolution operation, the feature of figure signal is obtained, by the figure signal
Tissue signature of the feature as the organizational structure, wherein the figure signal by each node in the relational network figure value group
At vector indicate.
Further, the figure convolution operation is indicated by carrying out Fourier transformation realization to figure signal are as follows:
Wherein, cGIndicate figure signal, UTIndicate the transposition of the eigenvectors matrix of Laplacian Matrix.
Further, the database sharing module is also used to logical according to the communication behavior feature construction of different time slice
Believe behavioral characteristic database;
The communication behavior feature construction that the relational network figure building module is also used to be sliced based on the different time closes
It is network;
The pass for the communication behavior feature construction that tissue signature's extraction module is also used to be sliced according to the different time
It is that network extracts tissue signature;
Correspondingly, the leaving office prediction module is also used to the corresponding tissue of communication behavior feature using different time slice
Feature and the method for machine learning are trained to obtain leaving office risk forecast model, the leaving office risk forecast model for carry out from
Duty risk profile.
The embodiment of the present invention utilizes the level information and communication behavior information architecture relational network figure of employee, and passes through figure letter
Number feature extraction, obtain the tissue signature of organizational structure, be trained using the method for the tissue signature and machine learning
To leaving office risk forecast model, the leaving office risk of employee in organizational structure is predicted by leaving office risk forecast model, without artificial
Intervene, provides objectively analysis prediction result based on artificial intelligence technology, manager is helped to find what organizational change generated early
Problem is simultaneously resolved.
Embodiment five
Fig. 5 is a kind of structural schematic diagram for computer equipment that the embodiment of the present invention five provides.Fig. 5, which is shown, to be suitable for being used to
Realize the block diagram of the exemplary computer device 512 of embodiment of the present invention.The computer equipment 512 that Fig. 5 is shown is only one
A example, should not function to the embodiment of the present invention and use scope bring any restrictions.
As shown in figure 5, computer equipment 512 is showed in the form of general purpose terminal.The component of computer equipment 512 can wrap
Include but be not limited to: one or more processor 516, storage device 528 connect different system components (including storage device 528
With processor 516) bus 518.
Bus 518 indicates one of a few class bus structures or a variety of, including storage device bus or storage device control
Device processed, peripheral bus, graphics acceleration port, processor or total using the local of any bus structures in a variety of bus structures
Line.For example, these architectures include but is not limited to industry standard architecture (Industry Subversive
Alliance, ISA) bus, microchannel architecture (Micro Channel Architecture, MAC) bus is enhanced
Isa bus, Video Electronics Standards Association (Video Electronics Standards Association, VESA) local are total
Line and peripheral component interconnection (Peripheral Component Interconnect, PCI) bus.
Computer equipment 512 typically comprises a variety of computer system readable media.These media can be it is any can
The usable medium accessed by computer equipment 512, including volatile and non-volatile media, moveable and immovable Jie
Matter.
Storage device 528 may include the computer system readable media of form of volatile memory, such as arbitrary access
Memory (Random Access Memory, RAM) 530 and/or cache memory 532.Computer equipment 512 can be into
One step includes other removable/nonremovable, volatile/non-volatile computer system storage mediums.Only as an example, it deposits
Storage system 534 can be used for reading and writing immovable, non-volatile magnetic media, and (Fig. 5 do not show, commonly referred to as " hard drive
Device ").Although being not shown in Fig. 5, the disk for reading and writing to removable non-volatile magnetic disk (such as " floppy disk ") can be provided and driven
Dynamic device, and to removable anonvolatile optical disk, such as CD-ROM (Compact Disc Read-Only Memory, CD-
ROM), digital video disk (Digital Video Disc-Read Only Memory, DVD-ROM) or other optical mediums) read-write
CD drive.In these cases, each driver can pass through one or more data media interfaces and bus 518
It is connected.Storage device 528 may include at least one program product, which has one group of (for example, at least one) program
Module, these program modules are configured to perform the function of various embodiments of the present invention.
Program/utility 540 with one group of (at least one) program module 542 can store in such as storage dress
It sets in 528, such program module 542 includes but is not limited to operating system, one or more application program, other program moulds
It may include the realization of network environment in block and program data, each of these examples or certain combination.Program module
542 usually execute function and/or method in embodiment described in the invention.
Computer equipment 512 can also be with one or more external equipments 514 (such as keyboard, sensing equipment, display
524 etc.) it communicates, the equipment interacted with the computer equipment 512 communication can be also enabled a user to one or more, and/or
(such as network interface card is adjusted with any equipment for enabling the computer equipment 512 to be communicated with one or more of the other calculating equipment
Modulator-demodulator etc.) communication.This communication can be carried out by input/output (I/O) interface 522.Also, computer equipment
512 can also by network adapter 520 and one or more network (such as local area network (Local Area Network,
LAN), wide area network (Wide Area Network, WAN) and/or public network, such as internet) communication.As shown in figure 5, net
Network adapter 520 is communicated by bus 518 with other modules of computer equipment 512.It should be understood that although not shown in the drawings,
Other hardware and/or software module can be used in conjunction with computer equipment 512, including but not limited to: microcode, device drives
Device, redundant processor, external disk drive array, disk array (Redundant Arrays of Independent
Disks, RAID) system, tape drive and data backup storage system etc..
The program that processor 516 is stored in storage device 528 by operation, thereby executing various function application and number
According to processing, such as realize leaving office Risk Forecast Method provided by the embodiment of the present invention, this method comprises:
Level information and communication behavior information architecture communication behavior property data base according to employee;
Relational network figure is constructed based on the communication behavior property data base, wherein the relational network figure is for describing
Communication behavior between the employee of different stage between the employee of same levels;
Feature extraction is carried out to the figure signal in the relational network figure, obtains the tissue signature of the organizational structure;
It is trained to obtain leaving office risk forecast model using the method for the tissue signature and machine learning, the leaving office wind
Dangerous prediction model is for carrying out leaving office risk profile.
Embodiment six
The embodiment of the present invention six additionally provides a kind of computer readable storage medium, is stored thereon with computer program, should
Risk Forecast Method of leaving office as provided by the embodiment of the present invention is realized when program is executed by processor, this method comprises:
Level information and communication behavior information architecture communication behavior property data base according to employee;
Relational network figure is constructed based on the communication behavior property data base, wherein the relational network figure is for describing
Communication behavior between the employee of different stage between the employee of same levels;
Feature extraction is carried out to the figure signal in the relational network figure, obtains the tissue signature of the organizational structure;
It is trained to obtain leaving office risk forecast model using the method for the tissue signature and machine learning, the leaving office wind
Dangerous prediction model is for carrying out leaving office risk profile.
The computer storage medium of the embodiment of the present invention, can be using any of one or more computer-readable media
Combination.Computer-readable medium can be computer-readable signal media or computer readable storage medium.It is computer-readable
Storage medium for example may be-but not limited to-the system of electricity, magnetic, optical, electromagnetic, infrared ray or semiconductor, device or
Device, or any above combination.The more specific example (non exhaustive list) of computer readable storage medium includes: tool
There are electrical connection, the portable computer diskette, hard disk, random access memory (RAM), read-only memory of one or more conducting wires
(ROM), erasable programmable read only memory (Erasable Programmable Read Only Memory, EPROM, or
Flash memory), optical fiber, portable compact disc read-only memory (CD-ROM), light storage device, magnetic memory device or above-mentioned
Any appropriate combination.In this document, computer readable storage medium can be any tangible Jie for including or store program
Matter, the program can be commanded execution system, device or device use or in connection.
Computer-readable signal media may include in a base band or as carrier wave a part propagate data-signal,
Wherein carry computer-readable program code.The data-signal of this propagation can take various forms, including but unlimited
In electromagnetic signal, optical signal or above-mentioned any appropriate combination.Computer-readable signal media can also be that computer can
Any computer-readable medium other than storage medium is read, which can send, propagates or transmit and be used for
By the use of instruction execution system, device or device or program in connection.
The program code for including on computer-readable medium can transmit with any suitable medium, including --- but it is unlimited
In wireless, electric wire, optical cable, radio frequency (Radio Frequency, RF) etc. or above-mentioned any appropriate combination.
The computer for executing operation of the present invention can be write with one or more programming languages or combinations thereof
Program code, described program design language include object oriented program language-such as Java, Smalltalk, C++,
It further include conventional procedural programming language-such as " C " language or similar programming language.Program code can be with
It fully executes, partly execute on the user computer on the user computer, being executed as an independent software package, portion
Divide and partially executes or executed on remote computer or terminal completely on the remote computer on the user computer.It is relating to
And in the situation of remote computer, remote computer can pass through the network of any kind --- including local area network (LAN) or extensively
Domain net (WAN)-be connected to subscriber computer, or, it may be connected to outer computer (such as provided using Internet service
Quotient is connected by internet).
Note that the above is only a better embodiment of the present invention and the applied technical principle.It will be appreciated by those skilled in the art that
The invention is not limited to the specific embodiments described herein, be able to carry out for a person skilled in the art it is various it is apparent variation,
It readjusts and substitutes without departing from protection scope of the present invention.Therefore, although being carried out by above embodiments to the present invention
It is described in further detail, but the present invention is not limited to the above embodiments only, without departing from the inventive concept, also
It may include more other equivalent embodiments, and the scope of the invention is determined by the scope of the appended claims.
Claims (14)
1. a kind of leaving office Risk Forecast Method characterized by comprising
Level information and communication behavior information architecture communication behavior property data base according to employee;
Relational network figure is constructed based on the communication behavior property data base, wherein the relational network figure is for describing difference
Communication behavior between the employee of rank between the employee of same levels;
Feature extraction is carried out to the figure signal in the relational network figure, obtains the tissue signature of the organizational structure;
It is trained to obtain leaving office risk forecast model using the method for the tissue signature and machine learning, the leaving office risk is pre-
Model is surveyed for carrying out leaving office risk profile.
2. the method according to claim 1, wherein the level information and communication behavior information according to employee
Construct communication behavior property data base, comprising:
Obtain the communication behavior information between organizational structure information, employee's essential information and employee, wherein the employee believes substantially
Breath includes tissue and employee's rank belonging to employee;
Screening statistics is carried out according to the communication behavior information between the organizational structure information, employee's essential information and employee, is obtained
The number that communication behavior occurs between the employee of rank between the employee of same rank is different to each employee;
The object and number of the rank of each employee and generation communication behavior are expressed as feature vector, and stores and arrives database
In, obtain communication behavior property data base.
3. the method according to claim 1, wherein the node in the relational network figure indicates communication behavior class
Type, the communication behavior type include in communication and each rank between the employee of any two different stage between employee
Communication, correspondingly, the communication time that the communication behavior type that the value of the relational network figure interior joint is indicated by the node occurs
Number is to indicate;
Side in the relational network figure indicates the simultaneous probability of different communication behavior type, correspondingly, the network of personal connections
In network figure while weight by possessing this simultaneously while corresponding two nodes represented by communication behavior type headcount come table
Show.
4. the method according to claim 1, wherein the figure signal in the relational network figure carries out spy
Sign is extracted, and the tissue signature of the organizational structure is obtained, comprising:
The relational network figure is analyzed using figure convolution operation, the feature of figure signal is obtained, by the feature of the figure signal
Tissue signature as the organizational structure, wherein the figure signal is made of the value of each node in the relational network figure
Vector indicates.
5. according to the method described in claim 4, it is characterized in that, the figure convolution operation is by carrying out in Fu to figure signal
Leaf transformation is realized, is indicated are as follows:
Wherein, cGIndicate figure signal, UTIndicate the transposition of the eigenvectors matrix of Laplacian Matrix.
6. -5 any method according to claim 1, which is characterized in that
The communication behavior feature of different time slice is stored in the communication behavior property data base;The relational network figure is
Communication behavior feature construction based on different time slice;The tissue signature is according to the logical of different time slice
Believe what the relational network figure of behavioural characteristic building extracted;
Correspondingly, the method using the tissue signature and machine learning is trained to obtain leaving office risk forecast model,
Include:
It is obtained using the corresponding tissue signature of communication behavior feature of different time slice and the method training of machine learning
Leaving office risk forecast model.
7. a kind of leaving office risk profile device characterized by comprising
Database sharing module, for the level information and communication behavior information architecture communication behavior characteristic according to employee
Library;
Relational network figure constructs module, for constructing relational network figure based on the communication behavior property data base, wherein described
Relational network figure is used to describe the communication behavior between the employee of different stage between the employee of same levels;
Tissue signature's extraction module obtains the tissue for carrying out feature extraction to the figure signal in the relational network figure
The tissue signature of framework;
Risk profile module obtains leaving office risk profile for being trained using the method for the tissue signature and machine learning
Model, the leaving office risk forecast model is for carrying out leaving office risk profile.
8. device according to claim 7, which is characterized in that the database sharing module includes:
Acquiring unit, for obtaining the communication behavior information between organizational structure information, employee's essential information and employee, wherein
Employee's essential information includes tissue and employee's rank belonging to employee;
Statistic unit is screened, for believing according to the communication behavior between the organizational structure information, employee's essential information and employee
Breath carries out screening statistics, obtains each employee and be different between the employee of rank to occur between the employee of same rank
The number of communication behavior;
Storage unit, for the object and number of the rank of each employee and generation communication behavior to be expressed as feature vector, and
It stores in database, obtains communication behavior property data base.
9. device according to claim 7, which is characterized in that the node in the relational network figure indicates communication behavior class
Type, the communication behavior type include in communication and each rank between the employee of any two different stage between employee
Communication, correspondingly, the communication time that the communication behavior type that the value of the relational network figure interior joint is indicated by the node occurs
Number is to indicate;
Side in the relational network figure indicates the simultaneous probability of different communication behavior type, correspondingly, the network of personal connections
In network figure while weight by possessing this simultaneously while corresponding two nodes represented by communication behavior type headcount come table
Show.
10. device according to claim 7, which is characterized in that tissue signature's extraction module is specifically used for:
The relational network figure is analyzed using figure convolution operation, the feature of figure signal is obtained, by the feature of the figure signal
Tissue signature as the organizational structure, wherein the figure signal is made of the value of each node in the relational network figure
Vector indicates.
11. device according to claim 10, which is characterized in that the figure convolution operation is by carrying out Fu to figure signal
In leaf transformation realize, indicate are as follows:
Wherein, cGIndicate figure signal, UTIndicate the transposition of the eigenvectors matrix of Laplacian Matrix.
12. according to any device of claim 7-11, which is characterized in that
The database sharing module is also used to the communication behavior feature construction communication behavior characteristic according to different time slice
According to library;
The communication behavior feature construction network of personal connections that the relational network figure building module is also used to be sliced based on the different time
Network figure;
The network of personal connections for the communication behavior feature construction that tissue signature's extraction module is also used to be sliced according to the different time
Network figure extracts tissue signature;
Correspondingly, the leaving office prediction module is also used to the corresponding tissue signature of communication behavior feature using different time slice
It is trained to obtain leaving office risk forecast model with the method for machine learning.
13. a kind of computer equipment characterized by comprising
One or more processors;
Storage device, for storing one or more programs,
When one or more of programs are executed by one or more of processors, so that one or more of processors are real
Now such as leaving office Risk Forecast Method as claimed in any one of claims 1 to 6.
14. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the program is by processor
Such as leaving office Risk Forecast Method as claimed in any one of claims 1 to 6 is realized when execution.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810325419.2A CN110378543A (en) | 2018-04-12 | 2018-04-12 | Leaving office Risk Forecast Method, device, computer equipment and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810325419.2A CN110378543A (en) | 2018-04-12 | 2018-04-12 | Leaving office Risk Forecast Method, device, computer equipment and storage medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110378543A true CN110378543A (en) | 2019-10-25 |
Family
ID=68243525
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810325419.2A Withdrawn CN110378543A (en) | 2018-04-12 | 2018-04-12 | Leaving office Risk Forecast Method, device, computer equipment and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110378543A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111340509A (en) * | 2020-05-22 | 2020-06-26 | 支付宝(杭州)信息技术有限公司 | False transaction identification method and device and electronic equipment |
CN111369044A (en) * | 2020-02-27 | 2020-07-03 | 腾讯云计算(北京)有限责任公司 | Method and device for estimating loss and computer readable storage medium |
CN111709714A (en) * | 2020-06-17 | 2020-09-25 | 腾讯云计算(北京)有限责任公司 | Method and device for predicting lost personnel based on artificial intelligence |
WO2021089013A1 (en) * | 2019-11-06 | 2021-05-14 | 中国科学院深圳先进技术研究院 | Spatial graph convolutional network training method, electronic device and storage medium |
-
2018
- 2018-04-12 CN CN201810325419.2A patent/CN110378543A/en not_active Withdrawn
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2021089013A1 (en) * | 2019-11-06 | 2021-05-14 | 中国科学院深圳先进技术研究院 | Spatial graph convolutional network training method, electronic device and storage medium |
CN111369044A (en) * | 2020-02-27 | 2020-07-03 | 腾讯云计算(北京)有限责任公司 | Method and device for estimating loss and computer readable storage medium |
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 |
CN111340509A (en) * | 2020-05-22 | 2020-06-26 | 支付宝(杭州)信息技术有限公司 | False transaction identification method and device and electronic equipment |
CN111340509B (en) * | 2020-05-22 | 2020-08-21 | 支付宝(杭州)信息技术有限公司 | False transaction identification method and device and electronic equipment |
CN111709714A (en) * | 2020-06-17 | 2020-09-25 | 腾讯云计算(北京)有限责任公司 | Method and device for predicting lost personnel based on artificial intelligence |
CN111709714B (en) * | 2020-06-17 | 2024-03-29 | 腾讯云计算(北京)有限责任公司 | Loss personnel prediction method and device based on artificial intelligence |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10963817B2 (en) | Training tree-based machine-learning modeling algorithms for predicting outputs and generating explanatory data | |
CN107330731B (en) | Method and device for identifying click abnormity of advertisement space | |
CN110378543A (en) | Leaving office Risk Forecast Method, device, computer equipment and storage medium | |
US20190272553A1 (en) | Predictive Modeling with Entity Representations Computed from Neural Network Models Simultaneously Trained on Multiple Tasks | |
US8862662B2 (en) | Determination of latent interactions in social networks | |
CN110175168A (en) | A kind of time series data complementing method and system based on generation confrontation network | |
CN110555047A (en) | Data processing method and electronic equipment | |
US20150161545A1 (en) | Visualization of spare parts inventory | |
US10402768B2 (en) | Business problem networking system and tool | |
CN113807520A (en) | Knowledge graph alignment model training method based on graph neural network | |
CN107507028A (en) | User preference determines method, apparatus, equipment and storage medium | |
US20210157819A1 (en) | Determining a collection of data visualizations | |
CN111368147A (en) | Graph feature processing method and device | |
CN112214775A (en) | Injection type attack method and device for graph data, medium and electronic equipment | |
CN105913235A (en) | Client account transfer relation analysis method and system | |
CN116089504B (en) | Relational form data generation method and system | |
CN110348581B (en) | User feature optimizing method, device, medium and electronic equipment in user feature group | |
Miller et al. | Post-glacial expansion dynamics, not refugial isolation, shaped the genetic structure of a migratory bird, the yellow warbler | |
US20230087204A1 (en) | Systems and methods to screen a predictive model for risks of the predictive model | |
CN115409541A (en) | Cigarette brand data processing method based on data blood relationship | |
CN114897607A (en) | Data processing method and device for product resources, electronic equipment and storage medium | |
CN110223125B (en) | User position obtaining method under node position kernel-edge profit algorithm | |
Najarzadeh | Testing equality of generalized variances of k multivariate normal populations | |
CN107402984B (en) | A kind of classification method and device based on theme | |
Porouhan | Optimization of overdraft application process with fluxicon disco |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20191025 |
|
WW01 | Invention patent application withdrawn after publication |