CN107291715A

CN107291715A - Resume appraisal procedure and device

Info

Publication number: CN107291715A
Application number: CN201610192971.XA
Authority: CN
Inventors: 王海英; 刘军宁; 郭鼎; 陈奥; 宋华青; 张晓莹; 石志伟; 金凯民; 袁泉; 刘长江
Original assignee: Alibaba Group Holding Ltd
Current assignee: Alibaba Group Holding Ltd
Priority date: 2016-03-30
Filing date: 2016-03-30
Publication date: 2017-10-24
Also published as: TW201741948A; WO2017167069A1

Abstract

The invention discloses a kind of resume appraisal procedure and device.Wherein, this method includes：Obtain history recruitment data acquisition system；Data are extracted from history recruitment data acquisition system, wherein, data at least include：One or more attributes corresponding recruitment result in position, one or more attributes are the parameters for being used in resume text data characterize applicant's feature, and recruitment result at least includes：Occurrence number, and/or one or more attribute of one or more attributes in position enroll number of times in position；The data being drawn into by training build resume assessment models；Resume assessment is carried out to the resume received using resume assessment models.The present invention solves in the prior art to be estimated applicant and carries out analysis comprehensively to evaluate applicant by the behavior to applicant, social data, and the behavior of applicant, social data are complicated and changeable and obtain that difficulty is high, so as to cause that evaluates to estimate the big technical problem of difficulty.

Description

Resume appraisal procedure and device

Technical field

The present invention relates to data processing field, in particular to a kind of resume appraisal procedure and device.

Background technology

In the recruitment of prior art, many recruitment websites by social networks, each side such as behavioral data can be obtained To the completeness of applicant as portraying, applicant is understood in all its bearings, so as to wish the help side of recruitment Find the suitable talent.For example, as " the seeking English net " of big data recruitment platform, with big data algorithm realize the talent with The accurate matching of enterprise's position, and with utility functions such as " key of position demand one are synchronized to main flow recruitment website "；" seek The characteristics of English net ", includes：(1) full dose data, form the three-dimensional user's job hunting image of various dimensions.Xun Ying Netcoms are too high Science and technology obtains the social information of job hunter, the speech that job hunter issues in forum and the paper delivered etc. and counted in all directions According to one people being extended into the photoplay of a people by past plane picture, so as to form the three-dimensional use of various dimensions Family job hunting image.(2) personalized, various dimensions dynamically analyze the rule of development of the talent and enterprise, and optimization bi-directional matching draws Hold up.English net is sought using the career development path of two million peoples as data source, by analysis, position is formed and promotes collection of illustrative plates, such as Position promotes path, position incidence relation etc..

The application similar with above-mentioned application function also includes " talent's radar ", and " talent's radar " is by everyone in network On the substantial amounts of data that leave, track of such as living, social words and deeds personal information, therefrom separate he interest graph, Personality is drawn a portrait and capability evaluation.

Therefore, existing program mainly carries out the matching of job applicant and position using social data, behavioral data etc., But there are the following problems for the matching of use social data, behavioral data progress job applicant and position：

(1) application target is excessively complicated

The most application scenarios of prior art concentrate on the matching that pairing is got married, foothold people attribute, in order to realize This target needs behavior to people, social data to analyze, and comprehensively people is evaluated and portrayed, so that Greatly promote the multifarious requirement of and data comprehensive to data.

(2) availability of data and comprehensive limited to

Prior art is to reach that the purpose for searching out suitable employee needs to collect many data, so to data Availability and the comprehensive of data have very high requirement, while also greatly having limited to the accuracy of technology.

Applied for for carrying out analysis comprehensively by the behavior to target person, social data in the prior art to evaluate target Person, causes to assess the problem of difficulty is big, effective solution is not yet proposed at present.

The content of the invention

The embodiments of the invention provide a kind of resume appraisal procedure and device, at least to solve in the prior art to applicant It is estimated and carries out analysis comprehensively to evaluate applicant by the behavior to applicant, social data, and the row of applicant It is complicated and changeable and obtain that difficulty is high for, social data, so as to cause that evaluates to estimate the big technical problem of difficulty.

One side according to embodiments of the present invention there is provided a kind of resume appraisal procedure, including：Obtain history recruitment Data acquisition system, wherein, history recruitment data acquisition system at least includes：Resume text data；Data acquisition system is recruited from history Middle extraction data, wherein, data at least include：One or more attributes corresponding recruitment result, one in position Or multiple attributes are the parameters for being used in resume text data characterize applicant's feature, recruitment result at least includes：One Or occurrence number, and/or one or more attribute of multiple attributes in position enroll number of times in position；Pass through training The data being drawn into build resume assessment models.

One side according to embodiments of the present invention, additionally provides a kind of resume appraisal procedure, including：Input is to be assessed Resume；The resume assessment result of resume to be assessed is obtained, wherein, resume assessment result is done according to resume assessment models Go out, resume assessment models extract data foundation according to from history recruitment data acquisition system.

Another aspect according to embodiments of the present invention, additionally provides a kind of resume apparatus for evaluating, including：Acquisition module, For obtaining history recruitment data acquisition system, wherein, history recruitment data acquisition system at least includes：Resume text data；Take out Modulus block, for extracting data from history recruitment data acquisition system, wherein, data at least include：One or more category Property in position corresponding recruitment result, one or more attributes are to be used in resume text data characterize applicant's feature Parameter, recruitment result at least include：Occurrence number of one or more attributes in position, and/or one or more Parameter enrolls number of times in position；Module is built, the data for being drawn into by training build resume assessment models； Evaluation module, for carrying out resume assessment to the resume received using resume assessment models.

One side according to embodiments of the present invention, additionally provides a kind of resume apparatus for evaluating, including：First input mould Block, for inputting resume to be assessed；Acquisition module, the resume assessment result for obtaining resume to be assessed, wherein, Resume assessment result is made according to resume assessment models, and resume assessment models are to recruit data acquisition system according to from history It is middle to extract what data were set up.

It is easily noted that, building for resume assessment models is carried out using the data extracted in history recruitment data acquisition system It is vertical, for building the priori knowledge to a certain position so that the analysis to applicant can be conceived to applicant's synthesis The excavation of the matching degree of strength and position, i.e. search out suitable resume for the different positions of different company, so The behavior to each applicant, social data can be removed to analyze, the complexity of recruitment is reduced, so as to remove from The tedious work of applicant's behavioral data in each social platform is collected, so as to further reduce in recruitment In the cost paid, effect and cost analogy face have more preferable performance.

Thus, the such scheme that the present invention is provided is solved to be estimated by applicant to applicant in the prior art Behavior, social data analysis comprehensively is carried out to evaluate applicant, and the behavior of applicant, social data are complicated and changeable And difficulty height is obtained, so as to cause that evaluates to estimate the big technical problem of difficulty.

Brief description of the drawings

Accompanying drawing described herein is used for providing a further understanding of the present invention, constitutes the part of the application, this hair Bright schematic description and description is used to explain the present invention, does not constitute inappropriate limitation of the present invention.In accompanying drawing In：

Fig. 1 is a kind of hardware block diagram of the terminal of according to embodiments of the present invention 1 resume appraisal procedure；

Fig. 2 is a kind of flow chart of according to embodiments of the present invention 1 optional resume appraisal procedure；

Fig. 3 is a kind of flow chart of according to embodiments of the present invention 1 optional resume appraisal procedure；

Fig. 4 is the flow chart of according to embodiments of the present invention 2 resume appraisal procedure；

Fig. 5 is a kind of schematic diagram of according to embodiments of the present invention 2 optional resume appraisal procedure；

Fig. 6 is a kind of structural representation of according to embodiments of the present invention 3 optional resume apparatus for evaluating；

Fig. 7 is a kind of structural representation of according to embodiments of the present invention 3 optional resume apparatus for evaluating；

Fig. 8 is a kind of structural representation of according to embodiments of the present invention 3 optional resume apparatus for evaluating；

Fig. 9 is a kind of structural representation of according to embodiments of the present invention 3 optional resume apparatus for evaluating；

Figure 10 is a kind of structural representation of according to embodiments of the present invention 3 optional resume apparatus for evaluating；

Figure 11 is a kind of structural representation of according to embodiments of the present invention 3 optional resume apparatus for evaluating；

Figure 12 is a kind of structural representation of according to embodiments of the present invention 3 optional resume apparatus for evaluating；

Figure 13 is a kind of structural representation of according to embodiments of the present invention 4 resume apparatus for evaluating；

Figure 14 is a kind of structural representation of according to embodiments of the present invention 4 optional resume apparatus for evaluating；

Figure 15 is a kind of structural representation of according to embodiments of the present invention 4 optional resume apparatus for evaluating；And

Figure 16 is a kind of structured flowchart of terminal according to embodiments of the present invention.

Embodiment

In order that those skilled in the art more fully understand the present invention program, below in conjunction with the embodiment of the present invention Accompanying drawing, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment The only embodiment of a present invention part, rather than whole embodiments.Based on the embodiment in the present invention, ability The every other embodiment that domain those of ordinary skill is obtained under the premise of creative work is not made, should all belong to The scope of protection of the invention.

It should be noted that term " first " in description and claims of this specification and above-mentioned accompanying drawing, " Two " etc. be for distinguishing similar object, without for describing specific order or precedence.It should be appreciated that this The data that sample is used can be exchanged in the appropriate case, so as to embodiments of the invention described herein can with except Here the order beyond those for illustrating or describing is implemented.In addition, term " comprising " and " having " and they Any deformation, it is intended that covering is non-exclusive to be included, for example, containing process, the side of series of steps or unit Method, system, product or equipment are not necessarily limited to those steps clearly listed or unit, but may include unclear It is that ground is listed or for the intrinsic other steps of these processes, method, product or equipment or unit.

Following embodiment may apply in common terminal, for example computer.Certainly the reality below Applying example can also be applied among server, and server is it also will be understood that equipment to be made up of one or more computers. Therefore, the structure of computer illustrated below is also applied for server.When mobile terminal computing capability progressively strengthens, Following examples can also be implemented in the terminal.Certainly, the step or module in following embodiments can divide Do not carried out in different server or terminal or mobile terminal, these servers or terminal or mobile terminal Between carry out necessary data interaction.

Embodiment 1

According to embodiments of the present invention, a kind of embodiment of resume appraisal procedure is additionally provided, it is necessary to illustrate, attached The step of flow of figure is illustrated can perform in the computer system of such as one group computer executable instructions, also, , in some cases, can be with different from order execution herein although showing logical order in flow charts Shown or described step.

The embodiment of the method that the embodiment of the present application 1 is provided can be in mobile terminal, terminal or similar fortune Calculate in device and perform.Exemplified by running on computer terminals, Fig. 1 is that a kind of resume of the embodiment of the present invention 1 is assessed The hardware block diagram of the terminal of method.As shown in figure 1, terminal 10 can include it is one or more (processor 102 can include but is not limited to Micro-processor MCV or can compile (one is only shown in figure) processor 102 Journey logical device FPGA etc. processing unit), the memory 104 for data storage and for communication function Transmitting device 106.It will appreciated by the skilled person that the structure shown in Fig. 1 is only signal, it is not right The structure of above-mentioned electronic installation causes to limit.For example, terminal 10 may also include it is more more than shown in Fig. 1 or Less component, or with the configuration different from shown in Fig. 1.

The resume that memory 104 can be used in the software program and module of storage application software, such as embodiment of the present invention Corresponding programmed instruction/the module of appraisal procedure, processor 102 is stored in the software program in memory 104 by operation And module, so as to perform various function application and data processing, that is, realize the Hole Detection of above-mentioned application program Method.Memory 104 may include high speed random access memory, may also include nonvolatile memory, such as one or many Individual magnetic storage device, flash memory or other non-volatile solid state memories.In some instances, memory 104 The memory remotely located relative to processor 102 can be further comprised, these remote memories can be connected by network It is connected to terminal 10.The example of above-mentioned network includes but is not limited to internet, intranet, LAN, shifting Dynamic communication network and combinations thereof.

Transmitting device 106 is used to data are received or sent via a network.Above-mentioned network instantiation may include The wireless network that the communication providerses of terminal 10 are provided.In an example, transmitting device 106 includes one Network adapter (Network Interface Controller, NIC), it can pass through base station and other network equipments It is connected to be communicated with internet.In an example, transmitting device 106 can be radio frequency (Radio Frequency, RF) module, it is used to wirelessly be communicated with internet.

Under above-mentioned running environment, this application provides resume appraisal procedure as shown in Figure 2.Fig. 2 is according to this hair The flow chart of the resume appraisal procedure of bright embodiment 1.

It should be noted that for foregoing each method embodiment, in order to be briefly described, therefore it is all expressed as to one it is The combination of actions of row, but those skilled in the art should know, the present invention is not limited by described sequence of movement System, because according to the present invention, some steps can be carried out sequentially or simultaneously using other.Secondly, art technology Personnel should also know that embodiment described in this description belongs to preferred embodiment, involved action and module Not necessarily necessary to the present invention.

Through the above description of the embodiments, those skilled in the art can be understood that according to above-mentioned implementation The method of example can add the mode of required general hardware platform to realize by software, naturally it is also possible to by hardware, but The former is more preferably embodiment in many cases.Based on it is such understand, technical scheme substantially or Say that the part contributed to prior art can be embodied in the form of software product, the computer software product is deposited Storage is in a storage medium (such as ROM/RAM, magnetic disc, CD), including some instructions are to cause a station terminal Equipment (can be mobile phone, computer, server, or network equipment etc.) is performed described in each embodiment of the invention Method.

Data acquisition system is used in following steps in the present embodiment, one or more data are carried out same or similar Processing, or may be considered a data acquisition system as the foundation of some action or step.

Step S21, obtains history recruitment data acquisition system, wherein, history recruitment data acquisition system at least includes：Resume Text data.

In above-mentioned steps, as optional embodiment, the source of history recruitment data acquisition system can be target recruitment The history recruitment data acquisition system of side, for example：Participate in trick in target recruitment side (as above recruitment season) in preset time The information of the personnel engaged, and participate in recruitment and obtain the information of the personnel of the target advertising unit position.Obtain history The method of recruitment data acquisition system can be that history recruitment data acquisition system is obtained by the database of website itself.

In a kind of optional embodiment, the history of nearly 5 years for obtaining target recruitment side by database is recruited data and made For the data acquisition system, due to the recruitment criterion of recruitment side can be over time change and change, for example, recruitment side is to learning The requirement gone through may be raised, or the working experience of applicant is more paid attention to, therefore, obtain whole history of nearly 2 years Recruitment data are used as data acquisition system as the part recruitment data of data acquisition system and the first five years to the first three years.

Step S23, data are extracted from history recruitment data acquisition system, wherein, the data of extraction at least include：One Individual or multiple attributes corresponding recruitment result in position, one or more attributes are to be used in resume text data characterize The parameter of applicant's feature；As an optional embodiment, recruitment result can include：One or more attributes Occurrence number, and/or one or more attributes in position enroll number of times in position.Certainly, recruitment result page can Including the other guide in addition to this two parts content, to be then conducive to the information of recruitment to count and recruit Engage in result.

In above-mentioned steps, the attribute of the above-mentioned parameter for characterizing applicant's feature can be the academic, special of applicant The parameters such as industry, working experience.

Step S25, the data being drawn into by training build resume assessment models.

In a kind of optional embodiment, the data being drawn into can be trained using GDBT algorithms, be built Resume assessment models, when being trained using GDBT algorithms to the data being drawn into, above-mentioned steps can be used first The history recruitment data acquisition system of extraction builds the decision tree in one or more dimensions, and the output result of final mask is many The accumulated value for the result that individual decision tree obtains.

Herein it should be noted that can be above-mentioned GDBT algorithms to the algorithm that the data being drawn into are trained, but not It is limited to this.The purpose that the data being drawn into are trained is the resume assessment models of structure is learnt the number being drawn into According to so that can obtain same or close defeated when resume assessment models receive same or analogous data again Go out result, therefore training needs substantial amounts of data, so that when being estimated using resume assessment models to resume, should The resume for the person of engaging can fall among the data for training, that is to say, that for the data volume of the data for training The application is not specifically limited, but in a preferred case, the data volume for being used to train being drawn into is bigger, several According to coverage it is wider, the resume assessment models order of accuarcy of structure is higher.

Step S27, optionally, the resume appraisal procedure that the above embodiments of the present application are provided can also comprise the following steps： Resume assessment is carried out to the resume received using resume assessment models.

In a kind of optional embodiment, the resume received is inputted into the resume model of structure, model is obtained Output result.In another optional embodiment, position expected from the output result of model and applicant can be entered Row matching, is pointed out if the match is successful.The mode of prompting also has many kinds, for example, can be by applicant's Different colors that name is used identify different matching degrees.

Herein it should be noted that due to the data that are extracted in step S23 include one or more attributes in position it is right The recruitment result answered, and one or more attributes are the parameters for being used in resume text data characterize applicant's feature, because This could be aware that, in step s 27 when the data to extraction are trained so as to build resume assessment models, not Using the behavior of recruitment person, social data, so as to avoid behavior, social data is complicated and changeable and obtains difficulty Gao Ying Ring the situation generation that resume assesses efficiency.It should also be noted that, the embodiment of the present application such scheme is using resume In data, these data easily get, and also use the parameter (attribute) of applicant and recruitment is tied The corresponding relation of fruit, in view of the data validity in resume is higher, therefore, not only will not because of without using behavior, Social data and cause the accuracy rate of assessment low, can also due to used history recruit result in applicant's parameter with recruitment As a result relation improves the degree of accuracy that resume is assessed.

It is easily noted that, by above-mentioned steps S21 to step S27, is taken out using in history recruitment data acquisition system The data taken carry out the foundation of resume assessment models, for building the priori knowledge to a certain position so as to applicant Analysis can be conceived to the excavation of the matching degree to applicant's comprehensive strength and position, i.e. for different company not Suitable resume is searched out with position, can so remove the behavior to each applicant, social data from and analyze, The complexity of recruitment is reduced, so that the tedious work for collecting applicant's behavioral data in each social platform is eliminated, So as to further reduce the cost paid in recruitment, effect and cost analogy face have more preferable performance.

Thus, the scheme that the above embodiment of the present invention is provided is solved to be estimated by right to applicant in the prior art The behavior of applicant, social data carry out analysis comprehensively to evaluate applicant, and the behavior of applicant, social data are multiple Miscellaneous changeable and acquisition difficulty height, so as to cause that evaluates to estimate the big technical problem of difficulty.

According to the above embodiments of the present application, data are extracted from history recruitment data acquisition system in step S23, data are extracted When can be extracted from whole history recruitment data acquisition systems, still, can be with as an optional embodiment Data are filtered or cleaned first, the data for thinking to be possible to produce influence are got rid of.For example, above-mentioned steps S23 may include steps of：

Step S231, is cleaned to history recruitment data acquisition system, wherein, the data cleansing is mainly used to assess not The resume passed through is shielded among history recruitment data acquisition system.

In above-mentioned steps, the purpose that history recruitment data acquisition system is cleaned is to find and history recruitment number is removed The noise data included according to set.

Step S233, data are extracted from the history recruitment data acquisition system after cleaning.

By above-mentioned two step, the data of extraction can be made more accurate, the resume assessment models of foundation is more accorded with Close and require.

May have many types for the resume that does not pass through of assessment, for example, assess the resume not passed through include it is following at least One of：

Due to staffing head count cause to assess not by resume, without carry out resume assessment and directly carry out Interview and interview unsanctioned resume, resume repetition delivery causes to assess the resume not passed through.

In above-mentioned steps, the human resources that head count can be used for characterizing recruitment side are directed to this in a certain position The current demand of position, the demand of future development and corporate comprehensive planning corporate integrated planning, the headcount predetermined to this position, And/or recruiter's quantity.

In a kind of optional embodiment, applicant's first meets the recruitment con dition of recruitment side, but due to Human Resource Department There is advance personnel amount planning for the position that applicant's first is applied for, if employing applicant's first, may draw The phenomenon of employee's redundancy is played, therefore does not employ applicant's first, the resume of applicant's first is due to establishment head count Cause to assess the resume not passed through.

In an alternative embodiment, due to any reason applicant second resume and without assessment, it is and direct Interview of the recruitment side in a certain position is take part in, but interview is not through, then and the resume of applicant's second is also considered as Assess the resume not passed through.

In another optional embodiment, applicant third reuses different sides to a certain position of application side repeatedly Formula delivers the resume of oneself, is thrown for example, being repeated several times by different recruitment websites to the same position of same recruitment side Resume is passed, the resume of applicant is also considered as resume not by assessment.

According to being actually needed for advertising unit the foundation of which type can certainly be set to be to assess the resume not passed through.

It is that the data being drawn into by training build resume assessment models according to the above embodiments of the present application, in step S25. All data being drawn into can be applied to and be set up in resume assessment models in this step, the resume assesses mould The inspection of type can use the resume truly received to test.This processing method needs to use true applicant Resume test, it is possible to by screening out for suitable applicant's mistake.As another optional embodiment, The data of extraction can be divided into two parts, a part is used for carrying out generating resume assessment models, and a part is used for life Into resume assessment models tested, in such an embodiment, step S25 may include steps of：

The data being drawn into are divided into training sample data and test sample data by step S251.

In above-mentioned steps, above-mentioned training sample data include multiple data, for building resume assessment mould by training Type, test sample data equally include multiple data, for verifying whether above-mentioned resume assessment models are accurate.

In a kind of optional embodiment, training sample data and test sample data can be history recruitment data acquisition systems In multiple dimensions on data.

Herein it should be noted that in order to ensure the accuracy of resume assessment models, being divided into training in the data of extraction It is diversified in any dimension that the data of sample data, which are, to ensure that a variety of data in any dimension can be through Cross training so that resume assessment models can learn a variety of data in any dimension.

Herein it should also be noted that, because training sample data are used to constitute resume assessment models, test sample data Order of accuarcy for examining resume assessment models, therefore, training sample data and test sample data are to have known The history recruitment data acquisition system of dawn recruitment result.

Step S253, is trained using training sample data and generates resume assessment models to be tested.Below with one Optional embodiment is illustrated.

In the optional embodiment, it is assumed that need recruitment in advance is database maintenance post, has been applied for successfully The personnel of the position have 6, first are used to train using the resume of this 6 personnel：

Personnel 1：A schools, A specialties, B companies, B positions；

Personnel 2：B schools, A specialties, C companies, B positions；

Personnel 3：A schools, B specialties, company A, A positions；

Personnel 4：C schools, B specialties, B companies, A positions；

Personnel 5：A schools, A specialties, C companies, B positions；

Personnel 6：C schools, B specialties, C companies, A positions.

It is possible thereby to obtain the feature of data maintenance hilllock personnel：

School：A schools occurred 3 times in 6 data, and accounting is 0.5；B schools occurred in 6 data 1 time, accounting is 0.17；C schools went out twice, and accounting is 0.33；Other schools did not occur, and accounting is 0.

Specialty：A specialties occurred 3 times, and accounting is 0.5；B specialties occurred 3 times, and accounting is 0.5；Other are special Industry did not occur, and accounting is 0.

Company：Company A occurred 1 time, and accounting is 0.17；B companies occurred 2 times, and accounting is 0.33；C is public Department occurred 3 times, and accounting is 0.5；Other companies did not occur, and accounting is 0.

Position:A positions occurred 3 times, and accounting is 0.5；B positions occurred 3 times, and accounting is 0.5；Other positions Accounting is 0.

In the resume of 6 personnel of the successful position has been applied for, there are 3 different schools, 3 are not Same company, and only there are 2 different specialties and position, therefore, for the angle of resume selection, specially Industry and position ratio school and company are more important, and its significance level is 1.5 times of school and company.Therefore, instructed School, company, specialty, the weight of position are respectively in experienced process：0.2、0.2、0.3、0.3.

By above-mentioned numerical value come the score of computing staff：

Personnel 1：A schools, A specialties, B companies, B positions；

Can calculate the score on each attribute of the personnel of obtaining 1 be respectively 0.5*0.2,0.3*0.5,0.2*0.33, 0.3*0.5, as 0.1,0.15,0.066 and 0.15；

Personnel 2：B schools, A specialties, C companies, B positions；

Can calculate the score on each attribute of the personnel of obtaining 2 be respectively 0.2*0.17,0.3*0.5,0.2*0.5 and 0.3*0.5, as 0.034,0.15,0.1 and 0.15；

Personnel 3：A schools, B specialties, company A, A positions；

Can calculate the score on each attribute of the personnel of obtaining 3 be respectively 0.2*0.5,0.3*0.5,0.2*0.17 and 0.3*0.5, as 0.1 0.15 0.034 0.15；

Personnel 4：C schools, B specialties, B companies, A positions；

It is respectively 0.2*0.33,0.3*0.5,0.2*0.33 that the score on each attribute of the personnel of obtaining 4, which can be calculated, And 0.3*0.5, as 0.066,0.15,0.066 and 0.15；

Personnel 5：A schools, A specialties, C companies, B positions

Can calculate the score on each attribute of the personnel of obtaining 5 be respectively 0.2*0.5,0.3*0.5,0.2*0.5 and 0.3*0.5, as 0.1,0.15,0.1 and 0.15；

Personnel 6：C schools, B specialties, C companies, A positions

Can calculate the score on each attribute of the personnel of obtaining 6 be respectively 0.2*0.33,0.3*0.5,0.2*0.5 and 0.3*0.5, as 0.066,0.15,0.1 and 0.15.

6 vectors are so can be obtained by, if using fairly simple algorithm, each personnel can be summed (when So, it would however also be possible to employ other modes calculate to enter row vector), the score 0.466 of personnel 1, the score 0.434 of personnel 2, The score 0.434 of personnel 3, the score 0.433 of personnel 4, the score 0.5 of personnel 5, the score 0.466 of personnel 6.Among these Score range is [0.433,0.5], thinks that this resume does not meet this position if less than 0.433.

Herein it should be noted that above-described embodiment only for the purpose of description, training data only has 6, when training number According to more extensive, the interval range of score will be more reasonable.

Above-mentioned is only the score of a position, can also make the score for drawing multiple positions in the same way, or many The score of multiple different positions in individual company, then, multiple attributes in job seeker resume is calculated, which fall into The scope of individual position, the then it is considered that personnel meet the requirement of the position.

Herein it should be noted that the method that above-described embodiment is provided can be used for being trained training sample data obtaining Resume assessment models, but obtain the method for resume assessment models and be not limited to that, it is any to pass through number of training Above-mentioned steps can be applied to according to the algorithm for obtaining resume assessment models, for example, GBDT algorithms etc..

Resume assessment models to be tested are tested, what is upchecked by step S255 using test sample data In the case of confirm resume assessment models to be tested be accurate resume assessment models.

After resume assessment models to be tested are obtained, because the value of training sample data might not be comprehensive, or instruction Practice the influence that there is the reasons such as noise data in sample data so that the degree of accuracy of resume assessment models to be tested is not It is necessarily higher, accordingly, it would be desirable to by be tested to verify to recruitment data model input test sample data to be tested Resume assessment models it is whether accurate.

Herein it should also be noted that, the data volume included in training sample data number with the standards of resume assessment models True degree is directly proportional.

By above-mentioned steps, existing data can be used to be verified come the resume assessment models to generation, so as to keep away The problem of possible mistake screens out resume caused by being verified using true recruitment data is exempted from.

For the ease of being calculated, vectorization processing can be carried out for above-mentioned data.That is, it is above-mentioned according to the application Embodiment, generating the training sample data of resume assessment models to be tested can be：Enter row vector to extract and to vector Data after extraction carry out the data obtained after feature arrangement；And/or, detect the test of resume assessment models to be tested Sample data can be：Enter the data after row vector is extracted and extracted to vector and carry out the data obtained after feature arrangement.

Training sample data and/or test sample data entered with row vector extract can be to training sample data and/or Data of the test sample data in one or more dimension are extracted.To training sample data and/or test specimens Notebook data enters to carry out feature to arrange being by shape in training sample data and/or test sample data after row vector is extracted The meaning identical data of the difference such as formula, form, display mode but sign are uniformly processed.The purpose of above-mentioned steps It is uniform data form, solves because same data have diversified forms caused by data variation, not easily pass through The technical problem that algorithm is trained.

According to the above embodiments of the present application, test sample data have been used to comment resume to be tested in above-mentioned steps S255 Estimate model to test, and confirm that resume assessment models to be tested are accurate resume assessment models.It is confirmed whether it is The mode of accurate resume assessment models has a variety of, is exported for example, actual result can will be considered to resume assessment models Result it is completely the same, just think accurate model.As another optional embodiment, step S255 can include：

Step S2551, test sample data input is tested into resume assessment models to be tested, and exports Assay.

Step S2553, if assay recruitment resultant error corresponding with test sample data is within a predetermined range, It is accurate resume assessment models then to confirm resume assessment models to be tested.

In above-mentioned steps, it is allowed to there is certain error, the presence of this error is probably the number due to extracting data Caused by amount is insufficient to, but the presence of the error has no effect on the judgement to recruiting result, also in acceptable model Within enclosing.

In a kind of optional embodiment, multiple data in test sample data are separately input into resume to be tested In assessment models, exemplified by being still database maintenance post by the post of above-mentioned recruitment, 3 in test sample data are Member through knowing to apply for result is tested：

Tester 1：It is A schools, C specialties, C companies, B positions, not admitted；

Tester 2：It is D schools, A specialties, B companies, A positions, not admitted；

Tester 3：B schools, A specialties, company A, A positions, successfully enroll.

Using the school obtained in above-described embodiment, company, specialty, the weight of position：0.2nd, 0.2,0.3,0.3, Calculate the test result of test sample data：

Tester 1：A schools, C specialties, C companies, B positions；

Can calculate obtain the score on each attribute of tester 1 be respectively 0.5*0.2,0.3*0,0.2*0.5, 0.3*0.5, as 0.1,0,0.1 and 0.15.

Tester 2：D schools, A specialties, B companies, A positions；

Can calculate obtain the score on each attribute of tester 2 be respectively 0*0.2,0.3*0.5,0.2*0.33, 0.3*0.5, as 0,0.15,0.066 and 0.15.

Tester 3：B schools, A specialties, D companies, A positions；

Can calculate obtain the score on each attribute of tester 3 be respectively 0.2*0.17,0.3*0.5,0.2*0, 0.3*0.5, as 0.034,0.15,0 and 0.15.

Score of the above-mentioned 3 test sample data on different dimensions is calculated, the score of tester 1 is obtained For 0.35, tester 2 is scored at 0.366, and tester 3 is scored at 0.434, only tester 3 Fall into the scoring span for meeting the position, therefore, the result obtained by above-mentioned resume assessment models is tester Member 1 is not admitted, and tester 2 is not admitted, and tester 3 is enrolled, identical with actual result, therefore can To think that above-mentioned resume assessment models have the higher degree of accuracy.

Herein it should be noted that when the result of test result and test sample data is differed, by the test sample Data are trained as training sample data to resume assessment models, until can obtain and test sample data As a result identical test result.

According to the above embodiments of the present application, entering row vector extraction includes：Step S2555. is to the one or more of applicant Attribute enters row vector extraction, wherein, one or more attributes include at least one of：Business Name, position title, School's title, major name；

Data after being extracted to vector, which carry out feature arrangement, to be included：One or more attributes are returned by step S2557 One change is handled.

Normalized processing mode has many kinds, in an optional embodiment according to Business Name, position title, School's title, the property of major name different there is provided several different normalization modes, this several normalization side Formula can be used individually, can also be used in combination.Normalized mode is not limited to this, other normalization sides Formula can also obtain identical effect.

Business Name normalized

Business Name is normalized including：Build industry vocabulary and ground noun list；According to industry vocabulary and ground Industry noun and place name noun in noun list extract Business Name, obtain the normalization result of Business Name.

In a kind of optional embodiment, due to Business Name generally by corporate site, exabyte, company's industry, The part of general term four is constituted, for example：Taobao (China) softcom limited, on this basis, constructs industry word Table and ground noun list are extracted to the title of company.

In an alternative embodiment, in Business Name by the case that place name and industry word are constituted, for example：In That parent company of state's architectural engineering etc..For the above-mentioned company extracted without obvious exabyte, place name and company's industry are extracted So that Business Name to be normalized.

In another optional embodiment, in addition it is also necessary to build the mapping vocabulary of company's English name and company's Chinese, And the mapping vocabulary of subsidiary and the parent company belonging to subsidiary, in the case where there is the company of English name, from Chinese corresponding with above-mentioned English name is searched in the mapping vocabulary of company's English name and company's Chinese, is obtained To the normalization result of Business Name, for example, Business Name is normalized into Alibaba for alibab Business Name； In the case of Business Name of the Business Name for subsidiary, in the mapping word of the parent company belonging to subsidiary and subsidiary The parent company belonging to above-mentioned subsidiary is searched in table, to obtain the normalized result of Business Name, for example, by exabyte The Business Name of referred to as Taobao is normalized to Alibaba.

Position title is normalized

Position title is normalized including：Confirm in history recruitment data acquisition system, occurrence number is more than pre- If the job description of number of times is correct position title；Job description in resume text data is built by editing distance Mapping vocabulary between correct position title；The job description in resume text data is existed by regular expression Matched in mapping vocabulary, the normalization result for the title that gets a duty.

In a kind of optional embodiment, in the example that above-mentioned preset times can be 2000 times, number of times will appear from big Accurate job description is confirmed as in the job descriptions of 2000 times, for example, in this post of application PHP Resume in, the number of times that the post is described as PHP is more than 2000, and the post is described as being soft Part design engineer, software engineer etc. then confirm the post for same post but occurrence number and not less than 2000 times Accurate entitled PHP, and be software design engineering teacher, software engineer by all job titles The title of post is normalized to PHP, if PHP and software design engineering teacher the two Job title is more than 2000, in the case where confirming that PHP and software design engineering teacher are same post, Confirm the title of the most entitled post of occurrence number.

In an alternative embodiment, by the duty that all job titles are software design engineering teacher, software engineer The title of business is normalized to PHP, is built first by editing distance in resume text data, software is set The title of this post of other characterization softwares Developmental Engineer such as meter engineer, software engineer and PHP Between mapping vocabulary, i.e. all titles for characterization software Developmental Engineer this post are both mapped into software Developmental Engineer this officially；By regular expression to the job description in resume text data mapping vocabulary in Matched, so as to obtain normalized result.

School's title is normalized

School's title is normalized including：By school's title in resume text data according to appearance herein According to order arrangement from large to small, and school's title of default ranking is obtained, obtain basic dictionary；To school's title Denoising is carried out, and school's title that denoising is obtained is matched by canonical carrying out with school's title in basic dictionary Matching, with school's title after being normalized；Using synonym table, school's title correspondence is constructed according to preset rules Write a Chinese character in simplified form, will appear from the corresponding title write a Chinese character in simplified form of school's title and be recorded as school's title；It is default by being included in school's title School's title of first suffix word is normalized to corresponding first suffix word, wherein, default first suffix word At least include：Vocationl Technical College, network institute, adult education, self-study examination and upgrading academic credential from junior college to university；Remove what is included in school's title Second suffix word, to generate school's title after normalization, wherein, the second suffix word is used for point for characterizing school Branch mechanism；Remove the prefix in school's title, to generate school's title after normalization.

In a kind of optional embodiment, by number of times of all school's titles occurred in resume text data according to appearance Arrange from high to low, and obtain school title formation base dictionary of the occurrence number ranking at first 1000, constituting base On the basis of plinth dictionary, denoising is carried out to school's title, after denoising is carried out to school's title, on basis Matched in dictionary with school title, to obtain the normalization result of school's title.

In an alternative embodiment, carrying out denoising to school's title is matched by canonical, discards school Data are caused in title, for example, karr Si Luer universities (Germany) are processed as into karr Si Luer universities；By tall building Door oceanography institute (it is that y trick systems are no that trick system is no is) is processed as Xiamen oceanography institute, and school's title is entered with abating the noise Influence during row normalized.

In another optional embodiment, school's title normalizing of default first suffix word will be included in school's title Corresponding first suffix word is turned to, for example, Xing Cai Vocationl Technical Colleges of Xiamen City are normalized into Vocationl Technical College； Remove the second suffix word for being used to characterize the branch of school included in school's title, for example, Jiangsu is scientific and technological School of Economics and Business Management of university is normalized to Jiangsu University of Science and Technology；Remove the prefix in school's title, for example, by Jiangsu Province Nanjing institute of Nanjing University is normalized to Nanjing institute of Nanjing University.

Herein it should be noted that the title table of elite school of the world can also be set up.Occur if there is any school's title Frequency is very low, then firstly the need of confirming whether school's title is elite school of the world, the method for confirmation can be in world's name School's title is searched in the title table in school.

Major name is normalized

Major name is normalized including：Special subject edition is built, and pattra leaves is carried out by special subject edition This model training；The specialty in resume text data is classified according to the Bayesian model after training.

Herein it should be noted that the JD texts and resume text of the resume text data to applicant in above-described embodiment This matching, the phase of method that can be based on tfidf and word2vec algorithms or based on the two algorithm evolution to text Quantified like property.

Fig. 3 shown in a kind of applying step of application scenarios, Fig. 3 be according to embodiments of the present invention 1 it is a kind of optional Resume appraisal procedure flow chart.As shown in figure 3, below under a kind of application scenarios of the above embodiments of the present application Example be described in detail it is as follows：

S31:Import history recruitment data acquisition system.

The history recruitment data acquisition system of one or more recruitment sides is obtained, above-mentioned historical data at least includes resume textual data According to.

S32:Data cleansing is carried out to history recruitment data acquisition system.

In above-mentioned steps, it can be from history recruits data acquisition system to carry out data cleansing to history recruitment data acquisition system The resume not passed through is assessed in shielding, wherein, it can be due to staffing head count to assess the resume not passed through Cause to assess not by resume, without carry out resume assessment and directly interviewed and interview unsanctioned resume, Resume, which repeats to deliver, causes to assess the resume not passed through.

S33:History recruitment data acquisition system is divided into training sample data and test sample data.

S34:Training sample data are carried out with vectorial extraction.

In above-mentioned steps, it can extract the application in training sample data to carry out vector to training sample data to extract One or more attributes of person, one or more attributes include one of following：Business Name, position title, school's name Title, major name.

S35:Feature arrangement is carried out to the vector of extraction.

In above-mentioned steps, it can be to the progress of one or more attributes of applicant to carry out feature whole to the vector of extraction Normalized.

S36:Model training is carried out to training sample data.

In above-mentioned steps, training sample data can be trained using GDBT algorithms, obtain resume to be tested Assessment models, but obtain the training algorithm not limited to this of resume assessment models.

S37:Feature arrangement is carried out to test sample data.

S38:By the way that training sample data are carried out with model training checking resume assessment models to be obtained.

S39:Output test result.

By test sample data input to resume assessment models to be tested, if the result of output and test sample data Recruitment result is identical, then it is considered that resume assessment models to be tested are accurate resume assessment models.

Embodiment 2

The embodiment of the present application additionally provides resume appraisal procedure as shown in Figure 4.Fig. 4 is according to embodiments of the present invention 2 Resume appraisal procedure flow chart.

Step S41, inputs resume to be assessed.

In above-mentioned steps, above-mentioned default resume assessment models can be any one letter in the embodiment of the present application 1 Go through assessment models.

Step S43, obtains the resume assessment result of resume to be assessed, wherein, resume assessment result is commented according to resume Estimate what model was made, resume assessment models extract data foundation according to from history recruitment data acquisition system.

Herein it should be noted that the default resume assessment models occurred in the above embodiments of the present application can be implemented Any one resume assessment models in example 1 or other resume assessment models in addition to embodiment 1, appoint The resume assessment models that meaning usage history recruitment data rather than social data are obtained can apply to the present embodiment.

It is easily noted that, building for resume assessment models is carried out using the data extracted in history recruitment data acquisition system It is vertical, for building the priori knowledge to a certain position so that the analysis to applicant can be conceived to applicant's synthesis The excavation of the matching degree of strength and position, i.e. search out suitable resume for the different positions of different company, so The behavior to each applicant, social data can be removed to analyze, the complexity of recruitment is reduced, so as to remove from The tedious work of applicant's behavioral data in each social platform is collected, so as to further reduce in recruitment In the cost paid, effect and cost analogy face have more preferable performance.Therefore, entered using above-mentioned resume assessment models Row resume assess, can scheme solve applicant is estimated by the behavior to applicant, social activity in the prior art Data carry out analysis comprehensively to evaluate applicant, and the behavior of applicant, social data are complicated and changeable and to obtain difficulty high, So as to cause that evaluates to estimate the big technical problem of difficulty.

In the above embodiments of the present application, step S43, the above method also includes：

Step S45, input history recruitment data acquisition system, wherein, the data in history recruitment set at least include：One Individual or multiple attributes corresponding recruitment result in position, one or more attributes are to be used in resume text data characterize The parameter of applicant's feature, recruitment result at least includes：Occurrence number of one or more attributes in position and/ Or one or more attributes enroll number of times in position.

In the above embodiments of the present application, step S43, after the resume assessment result of resume to be assessed is obtained, on Stating method also includes：

Step S451, the resume assessment result of object to be assessed is shown in predeterminable area.

In the application above-mentioned steps, after resume assessment result is obtained, it can be shown and built according to default display content Vertical assessment result.

It is the resume assessment models that are proposed in embodiment 1 to show using resume assessment models in a kind of optional embodiment Example, in user to default resume assessment system (wherein, the resume assessment system use default resume assessment models) Itself resume is inputted, or the form provided according to resume assessment system is filled in after itself resume, resume assessment system makes After being estimated with resume assessment models to the resume of user, the assessment result of above-mentioned user is obtained, and by the letter of user Assessment result is gone through to be shown on the display terminal shown in user, meanwhile, by the assessment to user's resume, resume is commented The position that user is adapted to can also be accessed by estimating system, therefore resume assessment system is in the resume assessment result of display user While, the suitable position of user's recommendation can also be shown as.

In an alternative embodiment, still using resume assessment models as embodiment 1 in the resume assessment models that propose For example, after manpower management personnel input the resume of one or more parts job hunter to default resume assessment system, letter Go through evaluation system to be estimated the resume of job hunter successively or by preset order, obtain assessment result and display and manpower On the display terminal of administrative staff, for example, can will meet position, not meet job hunter's list of position according to difference Color or position display, manpower management personnel can also by click on job hunter name or other operation to job hunter Specific resume checked.Meanwhile, resume assessment system can also be searched in resume inventory meets the position Job hunter is that manpower management personnel are recommended.

Herein it should be noted that the display mode for showing resume assessment result according to default display content is not limited to State any display mode of embodiment.

Fig. 5 is a kind of schematic diagram of optional resume appraisal procedure according to the embodiment of the present application 2, with reference to Fig. 5 Shown example, is provided above-described embodiment 2 on the basis of the resume appraisal procedure that the embodiment of the present application 1 is provided Method is further described.

Firstly, it is necessary to which explanation, this method can include two stages, first stage is the preproduction phase, that is, is taken Device acquisition history of being engaged in recruits data, and history recruitment data are trained using learning method, is assessed to obtain resume Model, the first stage is only carried out once in the case where history recruitment data do not change, or according to fixed cycle To update resume assessment models；Second stage is working stage, i.e., user is estimated using resume assessment models Stage, and this stage repeatability, repeatedly occur.

In a kind of optional embodiment, user inputs history to server by default resume assessment system and recruits number According to server is received after the history recruitment data of user's input, using machine learning method by recruiting data to history It is trained and obtains resume assessment models, said process can be the first stage of the application resume appraisal procedure.In clothes It is engaged in after device generation resume assessment models, user can use resume assessment models to be estimated resume, and the process can be with It is the second stage in the application resume appraisal procedure, user inputs new resume to be assessed to resume assessment models, Server is estimated to new resume, obtains the assessment result of new resume, and user's the reception server assesses what is obtained The assessment result of new resume, it is possible to which the assessment result to new resume carries out the operation such as showing.

Herein it should be noted that during the above-mentioned machine learning method in the embodiment can be the embodiment of the present application 1 Any one builds the method for resume assessment models.

Embodiment 3

According to embodiments of the present invention, a kind of resume apparatus for evaluating for being used to implement above-mentioned resume appraisal procedure is additionally provided, As shown in fig. 6, the device includes：Acquisition module 60, abstraction module 62, structure module 64 and evaluation module 66

Acquisition module 60 is used to obtain history recruitment data acquisition system, wherein, history recruitment data acquisition system at least includes：Letter Go through text data.

In the history recruitment data acquisition system that above-mentioned module is related to, the source of history recruitment data acquisition system can be that target is recruited The history recruitment data acquisition system for the side of engaging, for example：Target recruitment side in preset time (as above one recruitment season, or Upper two recruitment seasons) information of the personnel recruited is participated in, and participate in recruitment and obtain the target advertising unit position The information of personnel.The method for obtaining history recruitment data acquisition system can obtain history by the database of website itself Recruit data acquisition system.

In a kind of optional embodiment, the history of nearly 5 years for obtaining target recruitment side by database recruits data set Close, due to the recruitment criterion of recruitment side can be over time change and change, for example, recruitment side can to academic requirement It can raise, or the working experience of applicant is more paid attention to, therefore, obtain whole history recruitment data set of nearly 2 years Close and data are recruited in the first five years to the part of the first three years.

Abstraction module 62 is used to extract data from history recruitment data acquisition system, wherein, the data of extraction at least include： One or more attributes corresponding recruitment result in position, one or more attributes are to be used for table in resume text data Levy the parameter of applicant's feature；As an optional embodiment, recruitment result can include：One or more category Occurrence number, and/or one or more attribute of the property in position enroll number of times in position.Certainly, result page is recruited The other guide in addition to this two parts content can be included, being then conducive to the information of recruitment can count Recruit in result.

The data that building module 64 is used to be drawn into by training build resume assessment models.

Evaluation module 66 is used to carry out resume assessment to the resume received using resume assessment models.

Herein it should be noted that by above-mentioned module, being carried out using the data extracted in history recruitment data acquisition system The foundation of resume assessment models, for building the priori knowledge to a certain position so that the analysis to applicant can be reached Eye is in the excavation to applicant's comprehensive strength and the matching degree of position, and the different positions of as different company search out conjunction Suitable resume, can so remove the behavior to each applicant, social data from and analyze, and reduce the complexity of recruitment Degree, so as to eliminate the tedious work for collecting applicant's behavioral data in each social platform, is being recruited so as to reduce The cost paid during engaging, effect and cost analogy face have more preferable performance.

According to the above embodiments of the present application, with reference to shown in Fig. 7, above-mentioned abstraction module 62 can include：

Cleaning module 70, for history recruitment data acquisition system clean, wherein, the data cleansing be mainly used to by The resume not passed through is assessed to shield among history recruitment data acquisition system.

First extracts submodule 72, for extracting data from the history recruitment data acquisition system after cleaning.

According to the above embodiments of the present application, assessing the resume not passed through includes at least one of：Due to staffing Head count cause to assess not by resume, without resume assessment is carried out directly interviewed and interview difference Resume, resume repeats to deliver causes to assess the resume that does not pass through.For assessing the resume not passed through in embodiment 1 It is explained, will not be repeated here.

According to the above embodiments of the present application, with reference to shown in Fig. 8, above-mentioned structure module 64 is by training the number being drawn into According to structure resume assessment models.All data being drawn into can be applied in the module and set up resume assessment On model, the inspection of the resume assessment models can use the resume truly received to test.This processing side Method needs to use the resume of true applicant to test, it is possible to by screening out for suitable applicant's mistake.As The data of extraction, can be divided into two parts by another optional embodiment, and a part is commented for carrying out generation resume Estimate model, a part is used for testing the resume assessment models of generation, in such an embodiment above-mentioned structure module 44 can include：

Sort module 80, for the data being drawn into be divided into training sample data and test sample data.

Generation module 82, resume assessment models to be tested are generated for being trained using training sample data.

Specific example is substantially the same manner as Example 1, it is assumed that need recruitment in advance is database maintenance post, The personnel of the successful position of application have 6, first are used to train using the resume of this 6 personnel：

Personnel 1：A schools, A specialties, B companies, B positions；

Personnel 2：B schools, A specialties, C companies, B positions；

Personnel 3：A schools, B specialties, company A, A positions；

Personnel 4：C schools, B specialties, B companies, A positions；

Personnel 5：A schools, A specialties, C companies, B positions；

Personnel 6：C schools, B specialties, C companies, A positions.

By above-mentioned numerical value come the score of computing staff：

Personnel 1：A schools, A specialties, B companies, B positions；

Personnel 2：B schools, A specialties, C companies, B positions；

Personnel 3：A schools, B specialties, company A, A positions；

Personnel 4：C schools, B specialties, B companies, A positions；

Personnel 5：A schools, A specialties, C companies, B positions；

Personnel 6：C schools, B specialties, C companies, A positions；

6 vectors are so can be obtained by, if using fairly simple algorithm, can be summed to each personnel, people 1 score 0.466 of member, the score 0.434 of personnel 2, the score 0.434 of personnel 3, the score 0.433 of personnel 4, personnel 5 Score 0.5, the score 0.466 of personnel 6.Score range is [0.433,0.5] among these, is recognized if less than 0.433 This position is not met for this resume.

From unlike the example in embodiment 1, in the recruitment in this post, it is also contemplated that the age of personnel, If the age of personnel is within a predetermined range, fraction can be increased.For example, this post is that an experience is critically important Post, then, the age, more than 40 years old, can increase by 0.07 point.

Confirm module 74, for being tested using test sample data to resume assessment models to be tested, examining It is accurate resume assessment models that resume assessment models to be tested are confirmed in the case of.

According to the above embodiments of the present application, with reference to shown in Fig. 9, above-mentioned sort module 80 includes：

Second extracts submodule 90, is used to extract into row vector.

Sorting module 92, for carrying out feature arrangement to entering the data after row vector is extracted, to obtain training sample data And/or test sample data.

According to the above embodiments of the present application, with reference to shown in Figure 10, above-mentioned confirmation module 84 is for being confirmed.Confirm Whether it is that the modes of accurate resume assessment models has a variety of, for example, can will be considered to actual result assesses mould with resume The result of type output is completely the same, just thinks accurate model.It is used as another optional embodiment, above-mentioned confirmation mould Block 84 can include：

Inspection module 100, for test sample data input to be tested into resume assessment models to be tested, And export assay.

Submodule 102 is confirmed, if assay recruitment resultant error corresponding with test sample data is in preset range Interior, then it is accurate resume assessment models to confirm resume assessment models to be tested.

Tester 1：A schools, C specialties, C companies, B positions；

Tester 2：D schools, A specialties, B companies, A positions；

Tester 3：B schools, A specialties, D companies, A positions；

As another optional embodiment, when recruiting at that time, not the problem of not accounting for the age, and now such as The problem of fruit considers the age, tester 2 can be with bonus point, and now tester 2 is scored at 0.366+0.07=0.4336, in this manner it is achieved that tester 2 should be enrolled, but is not recorded really Take.It is within the acceptable range caused by recruitment needs change that this error, which is due to,.

According to the above embodiments of the present application, with reference to shown in Figure 11, above-mentioned second, which extracts submodule 90, is used for applicant One or more attributes enter row vector extraction, wherein, one or more attributes include at least one of：Exabyte Title, position title, school's title, major name；Above-mentioned sorting module 92 includes：Module 112 is normalized, is used for One or more attributes are normalized.

According to the above embodiments of the present application, with reference to shown in Figure 12, normalization module 112 can include at least one of：

First son normalization module 120, for building industry vocabulary and ground noun list；According to industry vocabulary and ground noun Industry noun and place name noun in table extract Business Name, the normalization result of Business Name are obtained, with to exabyte Title is normalized.

Second son normalization module 122, for confirming in history recruitment data acquisition system, occurrence number is more than default time Several job descriptions is correct position title；By the job description in editing distance structure resume text data and just Mapping vocabulary between true position title；By regular expression to the job description in resume text data in mapping Matched in vocabulary, the normalization result for the title that gets a duty, so that position title to be normalized.

3rd son normalization module 124, for by school's title in resume text data according to appearance herein according to Order arrangement from large to small, and school's title of default ranking is obtained, obtain basic dictionary；School's title is carried out Denoising, and school's title that denoising is obtained is matched by canonical matched with school's title in the dictionary of basis, With school's title after being normalized；Using synonym table, write a Chinese character in simplified form according to preset rules construction school title is corresponding, It will appear from the corresponding title write a Chinese character in simplified form of school's title and be recorded as school's title；Default first suffix will be included in school's title School's title of word is normalized to corresponding first suffix word, wherein, default first suffix word at least includes： Vocationl Technical College, network institute, adult education, self-study examination and upgrading academic credential from junior college to university；Remove the second suffix word included in school's title Language, to generate school's title after normalization, wherein, the second suffix word is used for the branch for characterizing school；Go Fall the prefix in school's title, to generate school's title after normalization, so that school's title to be normalized.

4th son normalization module 126, Bayes's mould is carried out for building special subject edition, and by special subject edition Type training；The specialty in resume text data is classified according to the Bayesian model after training, with to major name It is normalized.

It should be noted that above-mentioned normalized processing, has been carried out detailed description, herein in embodiment 1 Repeat no more.

Embodiment 4

According to embodiments of the present invention, a kind of resume for being used to implement the resume appraisal procedure in embodiment 2 is additionally provided to comment Estimate device, as shown in figure 13, the device includes：First input module 130 and acquisition module 132.

First input module 130, for inputting resume to be assessed；Acquisition module 132, for obtaining resume to be assessed Resume assessment result, wherein, resume assessment result is made according to resume assessment models, and resume assessment models are roots Set up according to extraction data in recruiting data acquisition system from history.

Herein it should be noted that the step that the acquisition module 132 of above-mentioned first input module 130 corresponds in embodiment 2 Rapid S41 to step S43, this module is identical with example and application scenarios that the step of correspondence is realized, but is not limited to State the disclosure of that of embodiment one.It should be noted that above-mentioned module may operate in reality as a part for device In the terminal 10 that the offer of example one is provided.

In the above embodiments of the present application, with reference to shown in Figure 14, said apparatus also includes：Second input module 140.

Second input module 140, for inputting history recruitment data acquisition system, wherein, in the history recruitment set Data at least include：One or more attributes corresponding recruitment result in position, one or more of attributes are institutes Stating is used for the parameter for characterizing applicant's feature in resume text data, the recruitment result at least includes：It is one or Occurrence number, and/or one or more of attribute of multiple attributes in the position enroll number of times in the position.

Herein it should be noted that the step S45 that above-mentioned second input module 140 corresponds in embodiment 2, this mould Block is identical with example and application scenarios that the step of correspondence is realized, but is not limited to the disclosure of that of above-described embodiment one. It should be noted that above-mentioned module may operate in the terminal 10 of the offer of embodiment one as a part for device In.

In the above embodiments of the present application, with reference to shown in Figure 15, said apparatus also includes：Display module 150, is used for The resume assessment result of the object to be assessed is shown in predeterminable area.

Herein it should be noted that above-mentioned display module 150 correspond to embodiment 2 in step S451, this module with The example that the step of correspondence is realized is identical with application scenarios, but is not limited to the disclosure of that of above-described embodiment one.Need It is noted that above-mentioned module may operate in the terminal 10 of the offer of embodiment one as a part for device In.

Embodiment 5

Embodiments of the invention can provide a kind of terminal, the terminal can be terminal group in Any one computer terminal.Optionally, in the present embodiment, above computer terminal can also be replaced with The terminal devices such as mobile terminal.

Optionally, in the present embodiment, above computer terminal can be located in multiple network equipments of computer network At least one network equipment.

In the present embodiment, above computer terminal can perform the program code of following steps in resume appraisal procedure： History recruitment data acquisition system is obtained, wherein, history recruitment data acquisition system at least includes：Resume text data；From history Data are extracted in recruitment data acquisition system, wherein, data at least include：One or more attributes are corresponding in position to recruit Engage result, one or more attributes are the parameters for being used in resume text data characterize applicant's feature, and recruitment result is extremely Include less：Occurrence number, and/or one or more parameter of one or more attributes in position are enrolled secondary in position Number；The data being drawn into by training build resume assessment models；The resume received is entered using resume assessment models Row resume is assessed.

Optionally, Figure 16 is a kind of structured flowchart of terminal according to embodiments of the present invention.As shown in figure 16, Terminal A can include：One or more (one is only shown in figure) processors 1601, memory 1603, And transmitting device 1605.

Wherein, the resume appraisal procedure that memory can be used in storage software program and module, such as embodiment of the present invention Programmed instruction/module corresponding with device, processor is stored in software program and module in memory by operation, So as to perform various function application and data processing, that is, realize above-mentioned resume appraisal procedure.Memory may include height Fast random access memory, can also include nonvolatile memory, such as one or more magnetic storage device, flash memory, Or other non-volatile solid state memories.In some instances, memory can further comprise remote relative to processor The memory that journey is set, these remote memories can pass through network connection to terminal A.The example of above-mentioned network includes But it is not limited to internet, intranet, LAN, mobile radio communication and combinations thereof.

Processor can call the information and application program of memory storage by transmitting device, to perform following step： History recruitment data acquisition system is obtained, wherein, history recruitment data acquisition system at least includes：Resume text data；From history Data are extracted in recruitment data acquisition system, wherein, data at least include：One or more attributes are corresponding in position to recruit Engage result, one or more attributes are the parameters for being used in resume text data characterize applicant's feature, and recruitment result is extremely Include less：Occurrence number, and/or one or more parameter of one or more attributes in position are enrolled secondary in position Number；The data being drawn into by training build resume assessment models；The resume received is entered using resume assessment models Row resume is assessed.

Optionally, above-mentioned processor can also carry out the program code of following steps：History recruitment data acquisition system is carried out Cleaning, wherein, data cleansing includes：The resume not passed through is assessed to shield among history recruitment data acquisition system：From clear Data are extracted in history recruitment data acquisition system after washing.

Optionally, above-mentioned processor can also carry out the program code of following steps：Due to staffing head count Cause to assess not by resume, without resume assessment is carried out directly interviewed and interview different resumes, letter Going through repetition and delivering causes to assess the resume not passed through.

Optionally, above-mentioned processor can also carry out the program code of following steps：The data being drawn into are divided into training Sample data and test sample data；It is trained using training sample data and generates resume assessment models to be tested； Resume assessment models to be tested are tested using test sample data, confirmed in the case where upchecking to be checked The resume assessment models tested are accurate resume assessment models.

Optionally, above-mentioned processor can also carry out the program code of following steps：Generation resume to be tested assesses mould The training sample data of type are：Enter the data after row vector is extracted and extracted to vector and carry out the number obtained after feature arrangement According to；And/or, detecting the test sample data of resume assessment models to be tested is：Enter row vector to extract and take out vector Data after taking carry out the data obtained after feature arrangement.

Optionally, above-mentioned processor can also carry out the program code of following steps：By test sample data input to treating Tested in the resume assessment models of inspection, and export assay；If assay and test sample data pair The recruitment resultant error answered confirms that resume assessment models to be tested assess mould for accurate resume within a predetermined range, then Type.

Optionally, above-mentioned processor can also carry out the program code of following steps：Entering row vector extraction includes：Correspondence The one or more attributes for the person of engaging enter row vector extraction, wherein, one or more attributes include at least one of：It is public Take charge of title, position title, school's title, major name；Data after being extracted to vector, which carry out feature arrangement, to be included： One or more attributes are normalized.

Optionally, above-mentioned processor can also carry out the program code of following steps：Business Name is normalized place Reason includes：Build industry vocabulary and ground noun list；According to the industry noun in industry vocabulary and ground noun list and place name name Word extracts Business Name, obtains the normalization result of Business Name；Position title is normalized including：Really Recognize in history recruitment data acquisition system, the job description that occurrence number is more than preset times is correct position title；It is logical The mapping vocabulary crossed between the job description in editing distance structure resume text data and correct position title；Pass through Regular expression is matched to the job description in resume text data in mapping vocabulary, and the title that gets a duty is returned One changes result；School's title is normalized including：By school's title in resume text data according to appearance Arranged herein according to order from large to small, and obtain school's title of default ranking, obtain basic dictionary；To learning School title carries out denoising, and matches the school's title for obtaining denoising and school's name in basic dictionary by canonical Title is matched, with school's title after being normalized；Using synonym table, school's name is constructed according to preset rules Title is corresponding to be write a Chinese character in simplified form, and be will appear from the corresponding title write a Chinese character in simplified form of school's title and is recorded as school's title；It will be wrapped in school's title School's title containing default first suffix word is normalized to corresponding first suffix word, wherein, after default first Sewing word at least includes：Vocationl Technical College, network institute, adult education, self-study examination and upgrading academic credential from junior college to university；Remove in school's title Comprising the second suffix word, with generate normalization after school's title, wherein, the second suffix word be used for characterize learn The branch in school；Remove the prefix in school's title, to generate school's title after normalization；Major name is entered Row normalized includes：Special subject edition is built, and Bayesian model training is carried out by special subject edition；According to Bayesian model after training is classified to the specialty in resume text data.

It will appreciated by the skilled person that terminal can also be smart mobile phone (such as Android phone, IOS mobile phones etc.), tablet personal computer, applause computer and mobile internet device (Mobile Internet Devices, MI D), the terminal device such as PAD.Figure 16 it does not cause to limit to the structure of above-mentioned electronic installation.For example, computer Terminal A may also include the component (such as network interface, display device) more or less than shown in Figure 16, or Person has the configuration different from shown in Figure 16.

One of ordinary skill in the art will appreciate that all or part of step in the various methods of above-described embodiment is can be with Completed by program come the device-dependent hardware of command terminal, the program can be stored in a computer-readable storage medium In matter, storage medium can include：Flash disk, read-only storage (Read-Only Memory, ROM), deposit at random Take device (Random Access Memory, RAM), disk or CD etc..

Embodiment 6

Embodiments of the invention additionally provide a kind of storage medium.Optionally, in the present embodiment, above-mentioned storage medium It can be used for preserving the program code performed by the resume appraisal procedure that above-described embodiment 1 is provided.

Optionally, in the present embodiment, above-mentioned storage medium can be located in computer network Computer terminal group In any one terminal, or in any one mobile terminal in mobile terminal group.

Optionally, in the present embodiment, storage medium is arranged to the program code that storage is used to perform following steps： History recruitment data acquisition system is obtained, wherein, history recruitment data acquisition system at least includes：Resume text data；From history Data are extracted in recruitment data acquisition system, wherein, data at least include：One or more attributes are corresponding in position to recruit Engage result, one or more attributes are the parameters for being used in resume text data characterize applicant's feature, and recruitment result is extremely Include less：Occurrence number, and/or one or more parameter of one or more attributes in position are enrolled secondary in position Number；The data being drawn into by training build resume assessment models；The resume received is entered using resume assessment models Row resume is assessed.

Optionally, above-mentioned storage medium is also configured to the program code that storage is used to perform following steps：Processor is also The program code of following steps can be performed：History recruitment data acquisition system is cleaned, wherein, data cleansing includes： The resume not passed through is assessed to shield among history recruitment data acquisition system：Taken out from the history recruitment data acquisition system after cleaning Access evidence.

Optionally, above-mentioned storage medium is also configured to the program code that storage is used to perform following steps：Due to personnel Establishment head count cause to assess not by resume, without resume assessment is carried out directly interviewed and interviewed Different resume, resume, which repeat to deliver, causes to assess the resume not passed through.

Optionally, above-mentioned storage medium is also configured to the program code that storage is used to perform following steps：It will be drawn into Data be divided into training sample data and test sample data；It is to be tested that generation is trained using training sample data Resume assessment models；Resume assessment models to be tested are tested using test sample data, what is upchecked In the case of confirm resume assessment models to be tested be accurate resume assessment models.

Optionally, above-mentioned storage medium is also configured to the program code that storage is used to perform following steps：Generation is to be checked The training sample data for the resume assessment models tested are：Enter the data after row vector is extracted and extracted to vector and carry out feature The data obtained after arrangement；And/or, detecting the test sample data of resume assessment models to be tested is：Enter row vector Data after extracting and being extracted to vector carry out the data obtained after feature arrangement.

Optionally, above-mentioned storage medium is also configured to the program code that storage is used to perform following steps：By test specimens Notebook data is inputted into resume assessment models to be tested and tested, and exports assay；If assay with The corresponding recruitment resultant error of test sample data confirms that resume assessment models to be tested are defined within a predetermined range, then True resume assessment models.

Optionally, above-mentioned storage medium is also configured to the program code that storage is used to perform following steps：Enter row vector Extraction includes：One or more attributes of applicant are entered with row vector to extract, wherein, one or more attributes include with It is at least one lower：Business Name, position title, school's title, major name；Data after being extracted to vector are carried out Feature, which is arranged, to be included：One or more attributes are normalized.

Optionally, above-mentioned storage medium is also configured to the program code that storage is used to perform following steps：To exabyte Title be normalized including：Build industry vocabulary and ground noun list；According to the row in industry vocabulary and ground noun list Industry noun and place name noun extract Business Name, obtain the normalization result of Business Name；Normalizing is carried out to position title Change processing includes：Confirm in history recruitment data acquisition system, the job description that occurrence number is more than preset times is correct Position title；By between the job description and correct position title in editing distance structure resume text data Map vocabulary；The job description in resume text data is matched in mapping vocabulary by regular expression, obtained To the normalization result of position title；School's title is normalized including：By in resume text data School title is arranged according to order from large to small herein according to appearance, and obtains school's title of default ranking, is obtained Basic dictionary；Denoising is carried out to school's title, and school's title that denoising is obtained and basis are matched by canonical School's title in dictionary is matched, with school's title after being normalized；Using synonym table, according to default Rule construct school title is corresponding to write a Chinese character in simplified form, and will appear from the corresponding title write a Chinese character in simplified form of school's title and is recorded as school's title； School's title comprising default first suffix word in school's title is normalized to corresponding first suffix word, wherein, Default first suffix word at least includes：Vocationl Technical College, network institute, adult education, self-study examination and upgrading academic credential from junior college to university；Go Fall the second suffix word included in school's title, to generate school's title after normalization, wherein, the second suffix word Pragmatic is in the branch for characterizing school；Remove the prefix in school's title, to generate school's title after normalization； Major name is normalized including：Special subject edition is built, and Bayes's mould is carried out by special subject edition Type training；The specialty in resume text data is classified according to the Bayesian model after training.

The embodiments of the present invention are for illustration only, and the quality of embodiment is not represented.

In the above embodiment of the present invention, the description to each embodiment all emphasizes particularly on different fields, and does not have in some embodiment The part of detailed description, may refer to the associated description of other embodiment.

, can be by other in several embodiments provided herein, it should be understood that disclosed technology contents Mode realize.Wherein, device embodiment described above is only schematical, such as division of described unit, It is only a kind of division of logic function, there can be other dividing mode when actually realizing, such as multiple units or component Another system can be combined or be desirably integrated into, or some features can be ignored, or do not perform.It is another, institute Display or the coupling each other discussed or direct-coupling or communication connection can be by some interfaces, unit or mould The INDIRECT COUPLING of block or communication connection, can be electrical or other forms.

The unit illustrated as separating component can be or may not be it is physically separate, it is aobvious as unit The part shown can be or may not be physical location, you can with positioned at a place, or can also be distributed to On multiple NEs.Some or all of unit therein can be selected to realize the present embodiment according to the actual needs The purpose of scheme.

In addition, each functional unit in each embodiment of the invention can be integrated in a processing unit, can also That unit is individually physically present, can also two or more units it is integrated in a unit.It is above-mentioned integrated Unit can both be realized in the form of hardware, it would however also be possible to employ the form of SFU software functional unit is realized.

If the integrated unit realized using in the form of SFU software functional unit and as independent production marketing or in use, It can be stored in a computer read/write memory medium.Understood based on such, technical scheme essence On all or part of the part that is contributed in other words to prior art or the technical scheme can be with software product Form is embodied, and the computer software product is stored in a storage medium, including some instructions are to cause one Platform computer equipment (can be personal computer, server or network equipment etc.) performs each embodiment institute of the invention State all or part of step of method.And foregoing storage medium includes：USB flash disk, read-only storage (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disc or CD Etc. it is various can be with the medium of store program codes.

Described above is only the preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art For member, under the premise without departing from the principles of the invention, some improvements and modifications can also be made, these improve and moistened Decorations also should be regarded as protection scope of the present invention.

Claims

1. a kind of resume appraisal procedure, it is characterised in that including：

History recruitment data acquisition system is obtained, wherein, the history recruitment data acquisition system at least includes：Resume text Data；

Data are extracted from history recruitment data acquisition system, wherein, the data at least include：One or many Individual attribute corresponding recruitment result in position, one or more of attributes are used in the resume text data In the parameter for characterizing applicant's feature, the recruitment result at least includes：One or more of attributes are described Occurrence number, and/or one or more of attributes in position enroll number of times in the position；

The data being drawn into by training build resume assessment models.

2. according to the method described in claim 1, it is characterised in that extracted from history recruitment data acquisition system described Data include：

History recruitment data acquisition system is cleaned, wherein, the data cleansing includes：Assessment does not pass through Resume from the history recruitment data acquisition system among shield：

The data are extracted from the history recruitment data acquisition system after cleaning.

3. method according to claim 2, it is characterised in that assessing the resume not passed through includes at least one of：

Due to staffing head count cause to assess not by resume, without carry out resume assessment and it is direct Interviewed and interview unsanctioned resume, resume repetition delivery causes to assess the resume not passed through.

4. according to the method described in claim 1, it is characterised in that the data being drawn into by training build resume and assessed Model includes：

The data being drawn into are divided into training sample data and test sample data；

It is trained using the training sample data and generates resume assessment models to be tested；

The resume assessment models to be tested are tested using the test sample data, upchecked In the case of confirm the resume assessment models to be tested be accurate resume assessment models.

5. method according to claim 4, it is characterised in that

Generating the training sample data of the resume assessment models to be tested is：Enter row vector to extract and to vector Data after extraction carry out the data obtained after feature arrangement；And/or,

Detecting the test sample data of the resume assessment models to be tested is：Enter row vector to extract and to vector Data after extraction carry out the data obtained after feature arrangement.

6. the method according to claim 4 or 5, it is characterised in that treated using the test sample data to described The resume assessment models of inspection are tested, and confirm that the resume assessment models to be tested are accurate resume Assessment models include：

The test sample data input is tested into the resume assessment models to be tested, and exported Assay；

If assay recruitment resultant error corresponding with the test sample data is within a predetermined range, It is accurate resume assessment models then to confirm the resume assessment models to be tested.

7. method according to claim 4, it is characterised in that

Entering row vector extraction includes：Row vector extraction is entered to one or more attributes of the applicant, wherein, One or more of attributes include at least one of：Business Name, position title, school's title, specialty Title；

Data after being extracted to vector, which carry out feature arrangement, to be included：One or more of attributes are normalized Processing.

8. method according to claim 7, it is characterised in that place is normalized to one or more of attributes Reason includes at least one of：

The Business Name is normalized including：Build industry vocabulary and ground noun list；According to described Industry noun and place name noun in industry vocabulary and described ground noun list extract the Business Name, obtain described The normalization result of Business Name；

The position title is normalized including：Confirm in history recruitment data acquisition system, go out The job description that occurrence number is more than preset times is correct position title；The resume is built by editing distance The mapping vocabulary between job description and the correct position title in text data；Pass through regular expression Job description in the resume text data is matched in the mapping vocabulary, the position name is obtained The normalization result of title；

School's title is normalized including：By school's title root in the resume text data According to being arranged according to order from large to small for appearance herein, and school's title of default ranking is obtained, obtain basis Dictionary；Carry out denoising to school's title, and by canonical match school's title for obtaining denoising with School's title in the basic dictionary is matched, with school's title after being normalized；Use synonym Table, construct that school title is corresponding to write a Chinese character in simplified form according to preset rules, will appear from that school's title is corresponding to write a Chinese character in simplified form Title is recorded as school's title；School's title of default first suffix word will be included in school's title Corresponding first suffix word is normalized to, wherein, the default first suffix word at least includes：Occupation Technical college, network institute, adult education, self-study examination and upgrading academic credential from junior college to university；Remove after second included in school's title Sew word, to generate school's title after normalization, wherein, the second suffix word is used to characterize The branch in school；Remove the prefix in school's title, to generate school's title after normalization；

The major name is normalized including：Special subject edition is built, and passes through the specialty point Class table carries out Bayesian model training；According to the Bayesian model after training in the resume text data Specialty classified.

9. a kind of resume appraisal procedure, it is characterised in that including：

Input resume to be assessed；

The resume assessment result of the resume to be assessed is obtained, wherein, the resume assessment result is according to resume What assessment models were made, the resume assessment models are to extract data foundation according to from history recruitment data acquisition system 's.

10. method according to claim 9, it is characterised in that methods described also includes：

History recruitment data acquisition system is inputted, wherein, the data in the history recruitment set at least include：One Or multiple attributes corresponding recruitment result in position, one or more of attributes are the resume text datas In be used to characterize the parameter of applicant's feature, the recruitment result at least includes：One or more of attributes exist Occurrence number, and/or one or more of attributes in the position enroll number of times in the position.

11. method according to claim 10, it is characterised in that assess knot in the resume for obtaining the resume to be assessed After fruit, methods described also includes：

The resume assessment result of the object to be assessed is shown in predeterminable area.

12. a kind of resume apparatus for evaluating, it is characterised in that including：

Acquisition module, for obtaining history recruitment data acquisition system, wherein, the history recruitment data acquisition system is at least Including：Resume text data；

Abstraction module, for extracting data from history recruitment data acquisition system, wherein, the data are at least Including：One or more attributes corresponding recruitment result in position, one or more of attributes are the letters Going through in text data is used for the parameter for characterizing applicant's feature, and the recruitment result at least includes：It is one or Occurrence number, and/or one or more of parameter of multiple attributes in the position are enrolled in the position Number of times；

Module is built, the data for being drawn into by training build resume assessment models.

13. device according to claim 12, it is characterised in that the abstraction module includes：

Cleaning module, for being cleaned to history recruitment data acquisition system, wherein, the data cleansing bag Include：The resume not passed through is assessed to shield among history recruitment data acquisition system：

First extracts submodule, for extracting the data from the history recruitment data acquisition system after cleaning.

14. device according to claim 13, it is characterised in that assessing the resume not passed through includes at least one of：

Due to staffing head count cause to assess not by resume, without carry out resume assessment and it is direct Interviewed and interview different resumes, resume repetition delivery causes to assess the resume not passed through.

15. device according to claim 12, it is characterised in that the structure module includes：

Sort module, for the data being drawn into be divided into training sample data and test sample data；

Generation module, resume assessment models to be tested are generated for being trained using the training sample data；

Module is confirmed, for being examined using the test sample data to the resume assessment models to be tested Test, it is accurate resume assessment models that the resume assessment models to be tested are confirmed in the case where upchecking.

16. device according to claim 15, it is characterised in that the sort module includes：

Second extracts submodule, is used to extract into row vector；

Sorting module, for carrying out feature arrangement to carrying out the data after the vector is extracted, to obtain the instruction Practice sample data and/or the test sample data.

17. the device according to claim 15 or 16, it is characterised in that the confirmation module includes：

Inspection module, for the test sample data input to be entered into the resume assessment models to be tested Performing check, and export assay；

Submodule is confirmed, if assay recruitment resultant error corresponding with the test sample data exists In preset range, then it is accurate resume assessment models to confirm the resume assessment models to be tested.

18. device according to claim 16, it is characterised in that

Described second extracts submodule, enters row vector for one or more attributes to the applicant and extracts, Wherein, one or more of attributes include at least one of：Business Name, position title, school's title, Major name；

The sorting module includes：Module is normalized, for place to be normalized to one or more of attributes Reason.

19. device according to claim 18, it is characterised in that the normalization module includes at least one of：

First son normalization module, for building industry vocabulary and ground noun list；According to the industry vocabulary and institute The industry noun and place name noun stated in ground noun list extract the Business Name, obtain returning for the Business Name One changes result, so that the Business Name to be normalized；

Second son normalization module, for confirming in history recruitment data acquisition system, occurrence number is more than pre- If the job description of number of times is correct position title；Built by editing distance in the resume text data Mapping vocabulary between job description and the correct position title；By regular expression to resume text Job description in notebook data is matched in the mapping vocabulary, obtains the normalization knot of the position title Really, so that the position title to be normalized；

3rd son normalization module, for by school's title in the resume text data according to appearance herein According to order arrangement from large to small, and school's title of default ranking is obtained, obtain basic dictionary；To described School's title carries out denoising, and matches the school's title for obtaining denoising and the basic dictionary by canonical In school's title matched, with school's title after being normalized；Using synonym table, according to default Rule construct school title is corresponding to write a Chinese character in simplified form, and will appear from the corresponding title write a Chinese character in simplified form of school's title and is recorded as institute State school's title；School's title comprising default first suffix word in school's title is normalized to accordingly The first suffix word, wherein, the default first suffix word at least includes：Vocationl Technical College, net Network institute, adult education, self-study examination and upgrading academic credential from junior college to university；Remove the second suffix word included in school's title, with life Into school's title after normalization, wherein, the second suffix word is used for the branch for characterizing the school； Remove the prefix in school's title, to generate school's title after normalization, to enter to school's title Row normalized；

4th son normalization module, pattra leaves is carried out for building special subject edition, and by the special subject edition This model training；The specialty in the resume text data is divided according to the Bayesian model after training Class, so that the major name to be normalized.

20. a kind of resume apparatus for evaluating, it is characterised in that including：

First input module, for inputting resume to be assessed；

Acquisition module, the resume assessment result for obtaining the resume to be assessed, wherein, the resume is assessed Result is made according to resume assessment models, and the resume assessment models are to recruit data acquisition system according to from history It is middle to extract what data were set up.

21. device according to claim 20, it is characterised in that described device also includes：

Second input module, for inputting history recruitment data acquisition system, wherein, in the history recruitment set Data at least include：One or more attributes corresponding recruitment result, one or more of attributes in position It is the parameter for being used in the resume text data characterize applicant's feature, the recruitment result at least includes：Institute Occurrence number, and/or one or more of attribute of one or more attributes in the position is stated to be reported Number of times is enrolled on position.

22. device according to claim 21, it is characterised in that described device also includes：

Display module, the resume assessment result for showing the object to be assessed in predeterminable area.