CN108009735A - A kind of resume appraisal procedure and device - Google Patents
A kind of resume appraisal procedure and device Download PDFInfo
- Publication number
- CN108009735A CN108009735A CN201711321692.XA CN201711321692A CN108009735A CN 108009735 A CN108009735 A CN 108009735A CN 201711321692 A CN201711321692 A CN 201711321692A CN 108009735 A CN108009735 A CN 108009735A
- Authority
- CN
- China
- Prior art keywords
- resume
- feedback
- feedback action
- assessed
- feature vector
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/10—Office automation; Time management
- G06Q10/105—Human resources
- G06Q10/1053—Employment or hiring
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Operations Research (AREA)
- Marketing (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Development Economics (AREA)
- Educational Administration (AREA)
- Data Mining & Analysis (AREA)
- Game Theory and Decision Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of resume appraisal procedure, including：Multiple first resumes corresponding with multiple first feedback data including the first feedback action item are obtained, negative sample is used as to each first resume construction feature vector；Multiple second resumes corresponding with multiple second feedback data including the second feedback action item are obtained, positive sample is used as to each second resume construction feature vector；Carry out the mapping relations between Training valuation model learning feature vector and the second feedback action item using positive sample and negative sample；And the probability that resume to be assessed is associated with the second feedback action in the case where being associated with the first feedback action is calculated according to mapping relations by feature vector of the assessment models based on resume to be assessed.The present invention discloses a kind of resume apparatus for evaluating.By the evaluation scheme of the embodiment of the present invention, relatively accurate assessment fraction can be provided to resume to be assessed using resume feedback data, strong technology auxiliary is provided for resume selection person.
Description
Technical field
The present invention relates to internet arena, more particularly to a kind of resume appraisal procedure and device.
Background technology
In recent years, as the development of Internet technology, the interconnection networking in personnel recruitment market become mainstream already.It is large-scale comprehensive
Close recruitment website (such as 51Job, the recruitment of intelligence connection) and the vertical recruitment website of internet industry (such as drag hook, Boss are directly engaged) covers greatly
Partial middle and high end job hunting crowd.In the case of job hunting colony resume common numbers, how to be lifted with technology help enterprise
The efficiency of personnel recruitment work is the important issue of recruitment website lifting industrial competition.
At present, some mature technologies of this respect mainly include：
1st, resume text automatically analyzes and information extraction, is taken out using natural language understanding technology from non-structured text
The key message of resume, such as educational background are taken, work unit, vocational skills, at the age, facilitate employment enterprise to be searched on website or be
Job requirement Auto-matching of the system for enterprise's issue；
2nd, the Auto-matching of resume and enterprise's job description：The job requirement that user's resume and enterprise are issued is made into letter respectively
Breath extracts and calculates matching degree, and the high resume of matching degree is sent to enterprise HR, can simplify the work of resume selection；
3rd, the resume search based on keyword and setting condition：The HR actives of employment enterprise are supported according to keyword and are preset
Condition (such as educational background, age, length of service, work city) potential candidate of search.
Although above-mentioned three types of technology both contributes to the efficiency of enterprise personnel recruitment, these technologies all focus on use
Natural language processing technique matches the information in each resume according to enterprises recruit persons for jobs demand, only plays the role of primary election, and
The immobilization of screening conditions can cause leakage sieve some suitable persons, resume selection person still need under many circumstances to resume into
Row is checked portionwise, thus the help for resume selection person is limited.
The content of the invention
In view of this, the embodiment of the present invention proposes a kind of resume appraisal procedure and device, it is possible to increase resume is carried out
The accuracy of assessment, effective reference is provided for resume selection person.
For this reason, the embodiment of the present invention proposes a kind of resume appraisal procedure, including：Obtain and include the first feedback action item
Corresponding multiple first resumes of multiple first feedback data, to each first resume construction feature vector be used as negative sample, institute
State the first feedback action item and represent that first resume is associated with the first feedback action；Obtain with including the second feedback action item
Corresponding multiple second resumes of multiple second feedback data, positive sample is used as to each second resume construction feature vector, described
Second feedback action item represents that second resume is associated with the second feedback action, and first feedback action is described second anti-
The feedback action of the previous stage of feedback action；Using the positive sample and negative sample come feature described in Training valuation model learning to
Mapping relations between amount and the second feedback action item；And pass through feature of the assessment models based on resume to be assessed
Vector calculates the resume to be assessed according to the mapping relations and the is associated with the case where being associated with the first feedback action
The probability of two feedback actions.
Preferably, the method further includes：The 3rd feedback data of the resume to be assessed is obtained, according to the described 3rd
Whether feedback data includes the probability of the second feedback action item and the resume to be assessed, corrects the assessment mould
The mapping relations that type is learnt.
Preferably, each first resume or the second resume construction feature vector are included：With first resume or
Two resumes and the correlation being associated between other resumes of second feedback action are simple to build first as one of feature
Go through or the feature vector of the second resume.
Preferably, first feedback action or the second feedback action for resume selection person for first resume or
Second resume actions taken.
Preferably, first feedback data and/or the second feedback data further include auxiliary information, the probabilistic correlation
In the auxiliary information.
Preferably, the auxiliary information includes the first job information, the first trade information and/or feedback time information.
Preferably, the method further includes：For the more parts of resumes to be assessed, based on each resume to be assessed
The probability is ranked up each resume to be assessed.
The embodiment of the present invention proposes a kind of resume apparatus for evaluating, including processor at the same time, and the processor operation is predetermined
Computer instruction to perform the resume appraisal procedure of any of the above-described embodiment.
By the embodiment of the present invention, it make use of first resume feedback data to establish assessment models, can pointedly learn
The mapping relations between feedback action of interest and the feature of resume are practised, for any resume to be assessed, assessment models energy
It is enough that relatively accurate assessment fraction is provided based on the mapping relations learnt according to the feature of the resume, provided for resume selection person
Strong technology auxiliary.
Brief description of the drawings
Fig. 1 is the indicative flowchart of one embodiment of the resume appraisal procedure of the present invention；
Fig. 2 is the indicative flowchart of another embodiment of the resume appraisal procedure of the present invention；
Fig. 3 is the indicative flowchart of one embodiment of the resume appraisal procedure of the present invention；
Fig. 4 is the indicative flowchart of another embodiment of the resume appraisal procedure of the present invention；
Fig. 5 is the indicative flowchart of one embodiment of the resume appraisal procedure of the present invention；
Fig. 6 is the indicative flowchart of another embodiment of the resume appraisal procedure of the present invention.
Embodiment
The multiple embodiments of the present invention are illustrated with reference to the accompanying drawings.
Fig. 1 is the indicative flowchart of one embodiment of the resume appraisal procedure of the present invention.
As shown in Figure 1, the resume appraisal procedure of the embodiment of the present invention includes：
S101, obtain multiple first resumes corresponding with multiple first feedback data including the first feedback action item；
Wherein, the first feedback action item represents that the first resume is associated with the first feedback action, and the first feedback data can be
Active or the feedback data automatically provided based on the first feedback action that resume selection person makes resume, such as resume selection
Person thinks to need to check details after the summary of resume is seen, then can perform check the action of resume details when or it
The first feedback data to the resume is actively provided afterwards, or resume selection person's execution is based on by resume data platform and checks resume
The action of details and automatically generate the first feedback data, in first feedback data can include represent check the first of details
Feedback action item.The job hunter that first feedback data can also be to provide resume is main after the feedback action of resume selection person is confirmed
It is dynamic to submit to the feedback data of resume data platform, for example, be invited to interview, by interview, by using phase etc., by resume number
Include representing such as inviting interview, the feedback data by interviewing feedback action item according to feedback generation of the platform based on job hunter.
First feedback action item is except that can be associated with checking resume details, invitation interview, be moved by feedbacks such as interviews
Outside the data item of work, the alternative qualified database of income, receipts for being associated with that such as resume selection person makes resume can also be
The feedback action of job hunter need not be notified by entering alternative interview storehouse etc..
It should be noted that for same resume, a feedback action item is generally comprised in its feedback data, when being subject to
When the resume of one stage feedback action is by feedback action of the choice of next stage to be associated with the next stage, the letter
The feedback action item for the last stage being associated with the feedback data gone through just is covered by the feedback action item of new stage, therefore resume
Feedback data item in feedback data can reflect the choice stage that the resume is presently in well.
In the embodiment of the present invention, when obtaining the first resume, come using feedback action of interest as the first feedback action
Corresponding first resume of the first feedback data for the first feedback action item for including being associated with first feedback action is obtained, thus
Acquired resume is the resume in the same choice stage, such as is similarly in the stage for being checked resume details, or place
In the stage for being invited to interview.
S102, be used as multiple positive samples to come Training valuation model learning institute the multiple first resume construction feature vector
State the mapping relations between feature vector and the first feedback action item；
, can be to these the first resumes after obtaining expression job hunter and being in multiple first resumes in same choice stage
In can be made of per a construction feature vector, feature vector the feature of multiple dimensions, some of which feature for example may be used
To be between correlation between the text of the first resume and industry characteristic, the text of the first resume and recruitment person's corporate culture
Correlation, the text of the first resume exist with the resume of other job hunters in same choice stage is hunted for a job and be in same industry
The text of correlation, the first resume on text is with hunting for a job same position in different industries and being in the same choice stage
The correlation of the resume of other job hunters on text etc..Here, the resume of other job hunters in the same choice stage by
The feedback data of other resumes identifies equally comprising the first feedback action item, i.e., it is anti-to be equally associated with first for other resumes
Present action item.As an example, when recruitment person's corporate culture or position demand be the job hunter of known certain overseas countries culture when,
It is just higher that job hunter's correlation in the experience of state's learning life is embodied in resume, then the value of this feature of resume is just
Can be higher；When between the text of the first resume and the resume of other job hunters in the same choice stage such as educational background,
When work experience etc. has more similarity, then the value of this feature of resume will be higher.
And so on, the feature that can largely explain the first feedback action can be built to the first resume.By by each
Feature Mapping is to fixed dimension, and using the value of feature as the value of this dimension, so for the first feedback coefficient of each
A feature vector can be constructed according to corresponding first resume, and based on associated first feedback of feedback action of interest
Action item and this feature vector is determined as positive sample, so can be obtained by multiple positive sample conducts for multiple first resumes
Training sample.
After having constructed training sample, a variety of machine learning models can be chosen as assessment models on these training samples
It is trained so that assessment models can learn between feature vector and the first feedback action item according to a large amount of training samples
Mapping relations.Machine learning model such as GBDT, Logistic Regression, Factorization Machine, DNN etc..
These models can acquire the mapping relations between feature vector and sample identification/attribute according to a large amount of training samples.
S103, by feature vector of the assessment models based on resume to be assessed calculate institute according to the mapping relations
State assessment fraction of the resume to be assessed on the first feedback action item.
After being trained to assessment models, any part resume to be assessed can be assessed with assessment models.Tool
For body, the feature vector of pattern same with training sample can be constructed to a resume to be assessed, and resume to be assessed
Feature vector is input in assessment models, assessment models can according to the mapping relations learnt, be resume to be assessed feature to
Each feature evaluation value in amount, so as to obtain assessment fraction of the resume to be assessed on the first feedback action item.
The embodiment of the present invention can be resume selection person by establishing assessment models respectively for different feedback action items
There is provided targetedly with reference to scoring, tackle the different demands of resume selection person, improve work efficiency.Such as check one for wishing more
The HR of a little candidate's resumes, can increasingly focus on the assessment models trained based on feedback action " checking details " to candidate's resume
The recommendation score provided, and for wishing the HR interviewed is unfolded as early as possible, it can increasingly focus on and face " is invited based on feedback action
The recommendation score that examination " or " passing through interview " provide candidate's resume and the assessment models of training.
By the embodiment of the present invention, it make use of first resume feedback data to establish assessment models, can pointedly learn
The mapping relations between feedback action of interest and the feature of resume are practised, for any resume to be assessed, assessment models energy
It is enough that relatively accurate assessment fraction is provided based on the mapping relations learnt according to the feature of the resume, so as to effectively utilize letter
Go through feedback data and strong technology auxiliary is provided for resume selection person, significantly reduce such as enterprise HR and line service supervisor and exist
Screen work load during resume.
Fig. 2 is the indicative flowchart of another embodiment of the resume appraisal procedure of the present invention.
As shown in Fig. 2, the method for the embodiment of the present invention includes：
S201, obtain multiple first resumes corresponding with multiple first feedback data including the first feedback action item, right
Multiple first resume construction feature vectors are used as multiple positive samples；
S202 obtains multiple second resumes corresponding with multiple second feedback data including the second feedback action item, to more
A second resume construction feature vector is used as multiple negative samples；
S203, come between Training valuation model learning feature vector and the first feedback action item using positive sample and negative sample
Mapping relations；
S204, calculated according to the mapping relations learnt by feature vector of the assessment models based on resume to be assessed and treated
Assess assessment fraction of the resume on the first feedback action item.
Step S201 and S204 in the embodiment of the present invention is with the corresponding steps in embodiment illustrated in fig. 1, herein only to step
Rapid S202 and S203 is illustrated.
Specifically, not only use to the feature vector of the first resume structure as positive sample, go back in the embodiment of the present invention
Use to the feature vector of the second resume structure as negative sample, to be trained jointly to assessment models with positive sample.This
In, it is the first feedback data in order to mutually be distinguished with the feedback data of the first resume that the second resume is corresponded to the second feedback data
Can be identical with the source of the second feedback data and form, simply the second feedback data fail to include the first feedback action item and
It is to include the second feedback action item, the second feedback action item represents that the second resume is associated with the second feedback action, and second is anti-
What feedback action represented is the feedback action of the previous either phase of the first feedback action.
Can be that positive sample assigns to enable assessment models to recognize positive sample and negative sample in the embodiment of the present invention
Value is 1 mark (label), and is the mark that negative sample assigns that value is 0.The value of mark is not limited to 0 and 1, hereinafter will
Illustrate.
As an example, when the first feedback action is invites interview, the second feedback action can be before inviting the interview stage
The feedback action of either phase, such as check resume summary or check resume details；When the first feedback action is to pass through interview,
Second feedback action can be by the feedback action of either phase before the interview stage, such as check resume summary, check resume
Details invite interview.In other words, the second resume be the failure to reach the first resume reach the stage in the previous choice stage
Do not considered by resume selection person or think to be not belonging to job hunter's resume of fit person after resume selection person considers.
By regarding the feature vector of these the second resumes as negative sample and as the first resume in the embodiment of the present invention
The positive sample of feature vector is together trained assessment models, and study schedule and the study that can accelerate assessment models are accurate
Property.
Can be also different in the second feedback action on the basis of embodiment illustrated in fig. 2 in some embodiments of the invention
The feedback action in stage sets different weighting coefficients, and adjusts negative sample according to the weighting coefficient of setting.For example, it is anti-to work as first
When feedback data item represents to pass through interview, in difference the second feedback data item of feedback action of different phase is represented, those tables
Show and only checked resume summary and do not checked that the feedback action items of the resume details reference value for assessment is relatively low, and
Those expressions have invited interview but have not had higher reference value for assessment by the feedback action item of interview, then may be used
Represent that its more important property is higher to set higher weighting coefficient to the second high feedback action of reference value, and to reference value
The second relatively low feedback action sets relatively low weighting coefficient and represents that its importance is relatively low, and is adjusted accordingly according to weighting coefficient
Negative sample, be, for example, that negative sample sets different important scale designation, or negative sample is replicated into more parts corresponding to weighting coefficient,
Etc..By the embodiment of the present invention, the study accuracy of assessment models can be further improved.
, can be after the 3rd feedback data of resume to be assessed be obtained, according to the in the various embodiments described above of the present invention
Whether three feedback data include the assessment fraction that the first feedback action item and assessment models have provided for resume to be assessed, repair
The mapping relations that positive assessment models are learnt.For example, assessment models are dynamic on some predetermined feedback to a resume to be assessed
Higher assessment fraction is provided as item, namely assessment result is to think that the resume to be assessed has higher probability to be subject to resume selection
During the respective feedback action of person, no matter whether the feedback action that resume selection person actually makes the resume to be assessed is assessment point
The targeted feedback action of number, can be first according to the 3rd feedback data and resume to be assessed that resume to be assessed accordingly obtains
Before obtained assessment fraction correct the mapping relations that assessment models are learnt, this is because the feature vector one to resume structure
As be made of many a features, it is difficult to which the value for each feature of a feature vector occur is scoring upper limit value or lower limit
Situation.The embodiment of the present invention can be assessment models by constantly learning to assessment result and actual feedback action
Splendid training sample is provided, further improves the study accuracy of assessment models, improves assessment degree of referring to.
In some embodiments of the invention, the first feedback data of the first resume is except the feedback including representing feedback action
Outside action item, other relevant auxiliary information auxiliary can also be included and assessed so that assessment models can be based at the same time
Auxiliary information provides assessment fraction to resume to be assessed, then given assessment fraction can be associated with auxiliary information, meets letter
Go through the different screening requirements of screening person.
In an embodiment of the invention, the auxiliary information in the first feedback data of the first resume can include the first duty
Position information.For example, when resume selection person needs to specific job applicant screening job hunter resume and (Ru Gehang during without considering specific industry
The occupations such as accounting that industry is required to, foreground), the first job information can be included in the feedback information of the first handled resume,
So that assessment models used training sample when learning mapping relations is all based on the job information, can pointedly expire
Sufficient resume selection person is to selecting the demand of the specific professional person.
In an embodiment of the invention, the auxiliary information in the first feedback data of the first resume can include the first row
Industry information and the first job information so that assessment models learnt used in training sample be all based on the spy of specific industry
Job orientation position, can pointedly help resume selection person to select the position fit person for possessing the sector experience.
In an embodiment of the invention, the auxiliary information in the first feedback data of the first resume can also include feedback
Temporal information, so when choosing the first resume, can be chosen according to time range of interest so that assessment models into
Training sample used in row study is all based on the feedback data in special time period, can pointedly help resume selection
Person selects candidate's resume in special time period.
It should be noted that, although the feedback data for being described above the first resume includes showing for a variety of auxiliary informations
Example, can not also include any auxiliary information in the embodiment of the present invention in the feedback data of the first resume, although or including
Auxiliary information but when choosing the first resume it is not intended that auxiliary information, it is possible thereby to assess the universality scoring of job hunter.Separately
On the one hand, it may also is that situation：The resume that resume data platform recommends resume selection person is adapted for example special
Job orientation position, therefore the first selected resume is actually to be associated with specific position, simply in its first feedback data simultaneously
Do not embody.
The embodiment of the present invention can be applied to several scenes.For example, for job hunter, resume data platform can be utilized to instruct
The assessment models perfected carry out the resume scoring for oneself；For resume selection person, then can utilize oneself, enterprise or resume data
The assessment models of platform training come for for example based on text matches select come more parts of candidate's resumes score, Ran Houke
With the sequence scored based on candidate's resume, candidate resume of the scoring more than certain threshold value is chosen from candidate's resume and carries out essence
Choosing.Resume selection person can also come with reference to other screening conditions such as candidate's work at present place, wages requirement, length of service etc.
Screening of candidates resume, and the selection result is ranked up according to the scoring of assessment models, the resume for ensureing to come above more accords with
Close the demand of resume selection person.So as to by the embodiment of the present invention, the work efficiency of resume selection person is greatly enhanced, dropped
Low work load, while improve the referring to property of resume recommendation.
Fig. 3 is the indicative flowchart of one embodiment of the resume appraisal procedure of the present invention.
As shown in figure 3, the resume appraisal procedure of the embodiment of the present invention includes：
S301, obtain multiple first resumes and its first feedback data；
S302 is to corresponding with the first feedback data including positive feedback action item a part of in multiple first resumes
One resume construction feature vector is used as multiple positive samples；
S303, to another part corresponding with the first feedback data including negative-feedback action item in multiple first resumes
First resume construction feature vector is used as multiple negative samples；
S304, come Training valuation model learning feature vector and positive feedback action using multiple positive samples and multiple negative samples
Mapping relations between；And
S305, by feature vector of the assessment models based on resume to be assessed calculate resume to be assessed according to mapping relations
Assessment fraction on positive feedback action item.
In the embodiment of the present invention shown in Fig. 3, S304-S305 is similar with the S203-S204 of embodiment illustrated in fig. 2, simply
Assessment models feedback action item of interest is different, and since feedback action item of interest is different, for the structure of training sample
It is also different to build mode, it is specifically described below.
In general, in embodiments of the present invention, it is not as paying close attention to like that in Fig. 1-embodiment illustrated in fig. 2 and is associated with resume
Some specific feedback action, but two classes will be divided into the associated all feedback actions of resume, one kind is moved for positive feedback
Make, one kind acts for negative-feedback.Wherein, positive feedback action can be checked including such as resume selection person to what the first resume was made
Resume details, invite and interview, approve that job hunter is the action of fit person in respective degrees by interview etc.；Negative-feedback acts
For such as resume selection person the first resume is made check resume summary action, due to but fail to cause resume selection person detailed
Go through the interest for seeing complete resume, it is believed that the condition of job hunter fails to comply with the minimum requirements of resume selection person.
Therefore, when the first resume is chosen in S301, can not have to according to whether including some in its first feedback data
Feedback action item determines whether it is satisfactory resume, but can obtain phase according to the required total amount of training sample
Answer the resume with feedback data of quantity, then in S302-S303 by these first resumes according to its first feedback data
Include represent positive feedback action positive feedback action item or represent negative-feedback action negative-feedback action item by this
A little first resumes are divided into the first resume of two parts, to wherein correspond to positive feedback action a part the first resume construction feature to
Amount is used as multiple positive samples, to wherein corresponding to the first resume of another part construction feature vector of negative-feedback action as multiple negative
Sample, then in S304 using these multiple positive samples and multiple negative samples come Training valuation model learning feature vector with it is positive and negative
The mapping relations between action item are presented, and are passed through in S305 for resume construction feature to be assessed vector input assessment models
Feature vector of the assessment models based on resume to be assessed calculates resume to be assessed on positive and negative according to the mapping relations learnt
Present the assessment fraction of action item.
Structure on the feature vector to resume and the training to assessment models are similar with previous embodiment, such as can
With related between other resumes of any positive feedback action to being associated with based on the first resume of positive feedback action is associated with
Property as one of feature build the feature vector of the first resume, omit illustrate herein.
In the embodiment of the present invention, since feedback action is divided into two classes in the structure of training sample, i.e. positive feedback is moved
Make and negative-feedback acts, thus the assessment models trained are that the study of mapping relations is carried out for all positive feedbacks action, energy
The feature for enough considering the associated resume of positive feedback action with different phase carries out comprehensive grading for resume to be assessed, so that
There is provided for resume selection person based on the reference value considered comprehensively.
, can also be such similar to previous embodiment in some embodiments of the invention, for different positive feedbacks action settings
Different weighting coefficients, and positive sample is adjusted according to the weighting coefficient of setting.For example, representing the positive feedback action of different phase
Different positive feedback data item in, the positive feedback action items of the resume details reference value for assessment has been checked in those expressions
It is relatively low, and interview has been invited in those expressions or has passed through the positive feedback action item interviewed for assessment with higher
Reference value, then positive feedback action that can be high to reference value set higher weighting coefficient and represent that its more important property is higher,
And the positive feedback action relatively low to reference value sets relatively low weighting coefficient and represents that its importance is relatively low, and can be according to weighting
Coefficient adjusts corresponding negative sample.By the embodiment of the present invention, the study accuracy of assessment models can be further improved.
On the utilization of weighting coefficient, if the machine learning model for supporting label to be real number is chosen, such as GBDT, can
Think that each positive sample sets the sample weights corresponding to weighting coefficient, such as be when the original of positive sample marks, the positive sample
When weighting coefficient is 5, the former mark 1 of the positive sample can be made to be multiplied by the 5 new mark as the positive sample obtained after weighting coefficient 5
Note.If choose and only support 0,1 machine learning model for mark, such as logistic regression (Logistic Regression), can
As soon as the number of the positive sample is replicated with the value according to weighting coefficient corresponding with a positive sample, for example weighting coefficient is incited somebody to action for 10
The positive sample replicates 10 times.Both approaches can allow assessment models more to pay attention to important positive feedback during study
Data, if following resume when resume is recommended is possible to obtain the feedback of importance higher, the assessment point of acquisition
Number will be higher.
Similarly, can also be after the second feedback data of resume to be assessed be obtained, according to second in the embodiment of the present invention
Whether feedback data has repaiied including positive feedback action item and assessment models for the assessment fraction provided of resume to be assessed
The mapping relations that positive assessment models are learnt.
Similarly, it can also include in the first feedback data of the first resume in the embodiment of the present invention or not include auxiliary letter
Breath, such as includes auxiliary information, then the assessment score correlations that assessment models provide are in auxiliary information.Here auxiliary information can also wrap
Include the first job information, the first trade information and/or feedback time information etc..
Similarly, the assessment models in the embodiment of the present invention can be used for more parts of candidate's resumes point to resume selection person
Assessment fraction is not provided, and the sequence that resume selection person can be scored based on candidate's resume, chooses scoring one from candidate's resume
Determine candidate's resume more than threshold value and carry out selected, the work efficiency of raising resume selection person.
Fig. 4 is the indicative flowchart of another embodiment of the resume appraisal procedure of the present invention.
As shown in figure 4, the resume appraisal procedure of the embodiment of the present invention includes：
S401, obtain multiple first resumes corresponding with multiple first feedback data including the first feedback action item, right
Each first resume construction feature vector is used as negative sample；
S402, obtain multiple second resumes corresponding with multiple second feedback data including the second feedback action item, right
Each second resume construction feature vector is used as positive sample；
S403, carry out using positive sample and negative sample Training valuation model learning described eigenvector and the described second feedback is dynamic
Make the mapping relations between item；
S404, by feature vector of the assessment models based on resume to be assessed calculate institute according to the mapping relations
State the probability that resume to be assessed is associated with the second feedback action in the case where being associated with the first feedback action.
Different from previous embodiment, it is anti-to represent that the first resume is associated with first for the first feedback action item in the embodiment of the present invention
Feedback acts, and the second feedback action item represents that the second resume is associated with the second feedback action, and the first feedback action is defined as second
The feedback action of the previous stage of feedback action, i.e. the first feedback action and the second feedback action are the feedback of two adjacent phases
Action, in other words, the embodiment of the present invention are concerned with the relation between the feedback action of two adjacent phases, or further and
Speech, it is contemplated that being subject to the possibility or general of the feedback action of next stage after the feedback action of previous stage is subject to
Rate, thus, the embodiment of the present invention is different from previous embodiment on the structure of training sample.
Specifically, in order to learn the relation between the feedback action of two adjacent phases, it is necessary first to be determined as pass
Gaze at target the latter half feedback action, such as it is contemplated that resume selection person for the invitation that resume is made to interview this anti-
Feedback act, then by invite interview be used as the second feedback action, using as invite interview previous stage check resume details this
One feedback action searches the resume with feedback data as the first feedback action, therefrom obtain with including check details this
Corresponding first resume of feedback data of one feedback action item and the feedback data pair that this feedback action item is interviewed including inviting
The second resume answered, to the feature vector that the first resume is built as negative sample, the feature vector conduct to the second resume structure
Positive sample, these positive samples and negative sample are trained assessment models as training sample, with learning characteristic vector with making
To pay close attention to the mapping relations between the second feedback action item of target.After training assessment models, for being already associated with first
The resume to be assessed of feedback action, can be obtained the resume to be assessed and is associated with second with construction feature vector input assessment models
The probability of feedback action.Here probability is also if commented for another angle in a kind of assessment fraction, such as the present embodiment
The scoring bound scope for estimating model is 0-100, when assessment models provide 60 points of scoring to resume, it is believed that should
The probability that resume is associated with the second feedback action on the basis of the first feedback action is associated with is 60%.
The training method of the building mode of the feature vector of resume and assessment models and foregoing implementation in the embodiment of the present invention
Example is similar, for example, using the first resume or the second resume be associated with the second feedback action other resumes between correlation as
One of feature builds the feature vector of the first resume or the second resume.Also there are one with previous embodiment for the embodiment of the present invention
Difference is the weighting coefficient without considering sample, without the resume feedback data for considering whole, such as when training is invited based on interview
Please this feedback action assessment models when, it is only necessary to considering, which includes expression, checks the feedback of this feedback data item of resume details
Data and and including represent invite interview this feedback data item feedback data both resumes, the former be used for design negative sample
This, the latter is used to design positive sample, and assessment models are trained accordingly.According to such setting, the assessment come is trained
Model can predict " in the case where resume selection person has checked resume details, send interview and invite " for resume to be assessed
Probability.
Therefore, can be according to resume selection person to the stage sexual act of candidate's resume and dynamically by the embodiment of the present invention
Provide the recommendation of next stage, as when from model recommend it is contemplated that check details candidate's resume in select predetermined quantity
After candidate's resume of interview is invited in preparation, model can provide corresponding job hunter respectively for resumes of these plan invitation interviews and lead to
The probability of interview is crossed, and can be based on sequence of the given probable value by the progress of these resumes from high to low, so as to be resume
Screening person provides accurate reference value, removes the work load that a part is further screened from, improves the work of resume selection person
Make efficiency, the situation of assessment models is slightly depended on suitable for resume selection person.
Similarly, can also be after the 3rd feedback data of resume to be assessed be obtained, according to the 3rd in the embodiment of the present invention
Whether feedback data, which includes, is used as concern the second feedback action item of target and previously to be assessed to this according to assessment models
What resume evaluated is associated with the probability of the second feedback action item, to correct the mapping relations that assessment models are learnt.
Similarly, it can also include in the first feedback data of the first resume in the embodiment of the present invention or not include auxiliary letter
Breath, such as includes auxiliary information, then the assessment probabilistic correlation that assessment models provide is in auxiliary information.Here auxiliary information can also wrap
Include the first job information, the first trade information and/or feedback time information etc..
Fig. 5 is the indicative flowchart of one embodiment of the resume appraisal procedure of the present invention, and the embodiment of the present invention is base
In the variant embodiment of embodiment illustrated in fig. 4.
As shown in figure 5, the appraisal procedure of the embodiment of the present invention includes：
S501, choose multigroup resume, every group of resume include with including a kind of multiple feedback data pair of feedback action item
A part of resume and another part resume corresponding with multiple feedback data including another feedback action item answered；
S502, for every group of resume, negative sample is used as to each construction feature of a part of resume vector, to another part resume
Each construction feature vector is used as positive sample；
S503, come using negative sample and positive sample Training valuation model learning feature vector and another feedback action item it
Between mapping relations；
S504, multiple assessment models based on multigroup resume establish fusion assessment models；
S505, by the fusion assessment models, the feature vector based on resume to be assessed calculates the letter to be assessed
Go through the probability for being associated with object feedback action item.
In embodiments of the present invention, the feedback action two for the multiple successive stages that resume may be made for resume selection person
Assessment models are built between two.For each group of resume, a kind of above-mentioned feedback action item represent feedback action be on
The feedback action of the previous stage for the feedback action that another feedback action item stated represents, and for multigroup resume and
Speech, have that the above-mentioned a kind of feedback action item in one group of resume represents be these successive stages feedback action in the initial period
Feedback action, separately have that the above-mentioned another feedback action item in one group of resume represents be these successive stages feedback action in
The feedback action of final stage, the feedback action of the final stage correspond to above-mentioned object feedback action item.Meanwhile except representing
Outside the feedback action item of the feedback action of initial period and the feedback action of expression final stage, the others of different resume groups
There is serial relation two-by-two between the feedback action that feedback action item represents.
In the embodiment of the present invention, in S501, when choosing every group of resume, two continuous ranks will be belonging respectively to including expression
Two kinds of resumes of the feedback action item of the feedback action of section are assigned in same group, so that a part of letter included in this group of resume
The feedback data gone through includes the feedback action in previous stage in two successive stages, another part letter that this group of resume includes
The feedback data gone through includes the feedback action of latter stage in the two successive stages.
In S502, for every group of resume, to the feedback action in previous stage that is associated with two successive stages
A part of resume construction feature vector is used as negative sample, the feedback action to the latter stage being associated with two successive stages
Another part resume construction feature vector be used as positive sample, to a part of resume or another part resume construction feature to
During amount, can by with a part of resume or another part resume respectively be associated with above-mentioned another feedback action item its
Correlation between his resume builds each feature vector as one of feature.
In S503, correspond to the assessment models of this group of resume with corresponding positive and negative sample training for every group of resume, learn
Mapping relations between the feedback action item of the feature vector for practising resume and the feedback action for representing the latter half.
Then, the key as the embodiment of the present invention is, it is necessary to will pass through the trained each assessment models of every group of resume
Certain algorithm is merged, and establishes fusion assessment models, and then resume to be assessed is carried out on mesh with fusion assessment models
Mark the probability assessment of feedback action.
For example, it can be directed to check that resume summary checks that details are assessed for the resume group training of the latter half feedback action
Model F1, interview assessment models F2 is invited for invite resume group of the interview for the latter half feedback action to train, and is directed to
To be trained by the resume group interviewed as the latter half feedback action by interviewing assessment models F3.Reference embodiment illustrated in fig. 4,
Each assessment models in these assessment models F1, F2 and F3, which can be independently operated, is respectively used to prediction for example to job orientation
In the case of position, resume selection person clicks in the case where seeing resume summary and checks the probability of resume complete information, resume
Screening person initiates the probability that interview is invited, and the probability that candidate passes through interview after resume details are checked.Shown in Fig. 5
In the embodiment of the present invention, after being merged to these three assessment models F1, F2 and F3, fusion assessment models F (F1 can be generated
(x), F2 (x), F3 (x)), wherein x is for the feature vector of resume structure.
In embodiments of the present invention, the overall acceptance probability conduct for candidate's resume can be provided to resume selection person
The overall recommendation of whole feedback stages, such as ought conform to the principle of simplicity to go through data platform or other approach recommend candidate's resume of predetermined quantity
Afterwards, fusion assessment models can directly give probability of the corresponding job hunter by interview to every part of candidate's resume, and can be based on
These resumes are carried out sequence from high to low by given probable value, are possessed so as to be provided for resume selection person compared with high reference
The suggested design of value, eliminates the most work burden of resume selection person, drastically increases the work of resume selection person
Make efficiency, the situation of assessment models is highly dependent on suitable for resume selection person.In fact, sample is trained through a large amount of priori in model
, with after the training of posteriority training sample, it is assessed accuracy and will be greatly improved, thus is relied on even for slight before for this
For the resume selection person of assessment models, it is also contemplated that height relies on assessment models to mitigate work load.
In embodiments of the present invention, to every part of resume in every group of resume can in a like fashion construction feature vector,
So that the feature vector of each resume includes the feature of identical dimensional.
In the embodiment of the present invention, by multiple assessment models merge into fusion assessment models amalgamation mode can have it is a variety of,
For example multiple assessment models are combined as fusion assessment models by the way that each assessment models are weighted with average mode, or pass through
Average mode is weighted after taking the logarithm to each assessment models again multiple assessment models are combined as fusion assessment models, etc.
Deng.
Similarly, can also be after the feedback data of resume to be assessed be obtained, according to the feedback coefficient in the embodiment of the present invention
According to whether including object feedback action item and according to the association that had previously been evaluated to the resume to be assessed of fusion assessment models
In the probability of object feedback action item, to correct the mapping relations that assessment models are learnt.
Similarly, it can also include or not wrap in the feedback data of each resume in the embodiment of the present invention in each group resume
Auxiliary information is included, such as includes auxiliary information, then the assessment probabilistic correlation that assessment models provide is in auxiliary information.Here auxiliary information
It can also include the first job information, the first trade information and/or feedback time information etc..
In addition, in embodiments of the present invention, screening person can also be included in the feedback data of each resume of each group resume
Information since the training sample of use is based on particular screen person and is built, so trained assessment models as auxiliary information
, can be using candidate letter of the higher accuracy to meet particular screen person's requirement by as the exclusive assessment models of particular screen person
Go through and provide higher assessed value, and the resume for the requirement of another screening person according to may obtain relatively low assessed value.
Below with reference to Fig. 6 to based on screening person's information and the scheme of Training valuation model is described in detail.
Fig. 6 is the indicative flowchart of another embodiment of the resume appraisal procedure of the present invention.
As shown in fig. 6, the resume appraisal procedure of the embodiment of the present invention can include：
S601, acquisition are corresponding with multiple first feedback data including the first feedback action item and first screening person's information
Multiple first resumes；
S602, be used as multiple positive samples to come Training valuation model learning institute the multiple first resume construction feature vector
State the mapping relations between feature vector and the first feedback action item；And
S603, by feature vector of the assessment models based on resume to be assessed calculate institute according to the mapping relations
State assessment fraction of the resume to be assessed on the first feedback action item.
This embodiment of the invention is the modification of embodiment illustrated in fig. 1.On the basis of embodiment illustrated in fig. 1, the present invention is real
First screening person's information that example further considers that the first feedback data of the first resume includes is applied, that is, the first feedback action
The first feedback action represented by is the action made by the first screening person to the first resume.In the embodiment of the present invention, except
Outside the resume for the feedback action that the selection of first resume makes resume based on same screening person, other processing procedures and Fig. 1
Illustrated embodiment is essentially identical.
It is different resumes according to the different demands of different resume selection persons or personal like by the embodiment of the present invention
Screening person trains different assessment models, these different assessment models may provide different recommendations to same candidate's resume
Value, being achieved in different resume selection persons can push away according to personalized demand to obtain for the personalized of candidate's resume
Recommend.
For example, in practical application, same position can have multidigit resume selection person sometimes, such as similarly for IOS high
Level engineer, may have the supervisor from two different departments to screen resume, and two people may have for the preference of candidate
Certain difference, the person of You Guo major companies resume is more prone to than being responsible for if any one.In theory, it can be this two resume selections
Person establishes a position respectively, but from the angle of enterprise HR, position division is carefully unfavorable for very much managing, so merging each business department
The demand of door is to being also common way in a position.And the embodiment of the present invention is due to the base when vectorial for resume construction feature
In the personal data of specific resume selection person, it is possible to realize the recommendation of personalization, that is to say, that check IOS for two
The operating officer of senior engineer's position candidate's resume, the assessment models of corresponding training will recommend different candidate's letters for it
Go through, this is that the existing resume based on text matches recommends method not reach.
In embodiments of the present invention, to the structure of the feature vector of the first resume, also there are some with above-mentioned other embodiment
Difference.Specifically, when to the first resume construction feature vector, it may be considered that with the first resume and be associated with same screening person
The first feedback action other resumes between correlation the feature vector of each first resume is built as one of feature, also
It is contemplated that correlation using between other resumes of the first feedback action of the first resume screening person different from being associated with as
One of feature builds the feature vector of each first resume.By related to screening person to consider during resume construction feature vector
Feature, the personal preference of particular screen person can be caught, consider the personal preference of screening person when scoring for candidate's resume
Inside, the accurate of personalization is provided for specific resume selection person to recommend.
In an embodiment of the invention, with previous embodiment similarly, can also be created for the training of assessment models negative
Sample.Such as it can obtain corresponding with multiple second feedback data including the second feedback action item and first screening person's information
Multiple second resumes, the plurality of second resume construction feature vector is carried out and positive sample together Training valuation as multiple negative samples
Model learning mapping relations.Wherein, the second feedback action item represents that the second resume is associated with the second feedback action, and the second feedback is dynamic
Work can be that feedback action (embodiment shown in Figure 4), the first feedback action of the previous stage of the first feedback action are previous
The feedback action (embodiment shown in Figure 2) in each stage or can also be the anti-of attribute opposite with the first feedback action
Feedback action (such as the first feedback action and the second feedback action are respectively positive and negative feedback action, referring specifically to embodiment illustrated in fig. 3).
Similarly, can also be after the 3rd feedback data of resume to be assessed be obtained, according to the 3rd in the embodiment of the present invention
Feedback data whether correct by the assessed value that has provided for resume to be assessed including the first feedback action item and assessment models
The mapping relations that assessment models are learnt.
Similarly, it can also include in the first feedback data of the first resume in the embodiment of the present invention or not include auxiliary letter
Breath, such as includes auxiliary information, then the assessment score correlations that assessment models provide are in auxiliary information.Here auxiliary information can also wrap
Include the first job information, the first trade information and/or feedback time information etc..
Similarly, the assessment models in the embodiment of the present invention can be used for more parts of candidate's resumes point to resume selection person
Do not provide assessment fraction or the assessment assessed value such as probability, resume selection person can the sequence based on candidate's resume assessed value, from time
Select the candidate's resume chosen and scored more than certain threshold value in resume to carry out selected, improve the work efficiency of resume selection person.
System is recruited to an exemplary cloud as the feedback data acquisition modes of resume in the embodiment of the present invention below
Illustrate, it should be noted that example that the present invention is not limited thereto, other feedback data acquisition modes such as job hunter or active
Submission feedbacks data to resume data platform or resume selection person owned enterprise and builds resume inventory simultaneously in resume data platform
Voluntarily record feedback action data or voluntarily build resume inventory after downloading resume data and voluntarily record feedback action data
It is also within the scope of the invention.
Different with traditional enterprises recruitment system, cloud recruitment system is a kind of SAAS (Software as
Service) platform.This platform is developed and safeguarded by third party, and can access internet has recruitment regulatory requirement to be all
Enterprises service.As user, using the enterprise of the platform without purchase, the system is disposed and safeguarded, it is only necessary to according to use
Number and usage time payment.This kind of cloud recruitment system has two kinds of modes for establishing and safeguarding personnel resume storehouse, and first
Kind situation is that such cloud recruitment system is developed by Large-Scale Interconnected net recruitment website in itself, can directly share the people of its parent company
Ability resume resource；The second situation is independent cloud recruitment system, it is necessary to which the enterprise side mandate system used uses enterprise's account
Number download Large-Scale Interconnected net recruitment website on enterprise issue the relevant all resumes of position.In the latter case, once cloud
Recruitment system possesses enough enterprise customers, can also build up the more comprehensive personnel resume storehouse of a ratio.
Enterprise customer newly can issue a position demand in cloud recruitment system or new use is sent for existing position
People applies, needs to provide post title for a position, and the length of service requires, and educational requirement, region requires, skill set requirements,
Responsibilities etc. describes, and also to provide the registration mailbox or system identifier of one or more resume selection persons.For same position
It may consider multiple resume selection persons.
For resume selection person when browsing the personnel resume of system recommendation, which resume simply simply skims through summary, which
A little resumes are clicked and detailed examination, which resume are marked as that interview can be invited, and all can recruit system by cloud records automatically
Get off.Subsequent candidate people participated in interview after, resume selection person also need to fill in systems to candidate interview evaluation and
Marking.Here, resume browses situation, detailed examination data knead dough examination invitation data, and traditional internet recruitment website also may be used
To obtain, but specific resume selection people generally can not be accurately corresponded to, enterprise account can only be associated with.And candidate interviewing
It is then that conventional internet recruitment website can not obtain to assess information.
Further, in cloud recruitment system, it is by whether receiving offer after interview and entering trade-after for candidate
No to have passed through the trial period, resume selection person can also record in systems.
Some data instances of system record are as follows：
<uid：xiaoming@Atech.com,position_id:45678,cv_id:23653490012,action:
go_detail,time_stamp:1510578275>
<uid：limei@Atech.com,position_id:45678,cv_id:23653490012,action:
invitation,time_stamp:1510576491>
<uid：xiaoming@byteScience.com,position_id:31682, cv_id:23653490012,
action:invitation,time_stamp:1510590021>
Here uid have recorded enterprise's mailbox of resume selection person (for one of screening person's information in above-described embodiment
Example), position_id have recorded the ID of specific position, and cv_id have recorded personnel resume ID, action and have recorded resume sieve
The action of the person of choosing, including check details (go_detail), invite the action such as interview (invitation), time_stamp records
The timestamp that action occurs.
, can be according to the side of any of the above-described embodiment after being collected into the resume feedback data of cloud recruitment platform user generation
Case, corresponding resume assessment models are trained according to these data.
The embodiment of the present invention additionally provides resume apparatus for evaluating, its realization is such as computer installation including processor,
The computer installation further includes memory, wherein computer executable instructions can be stored with, can design corresponding meter as needed
Calculation machine executable instruction stores in memory so that processor is able to carry out this hair when running the corresponding computer instruction
Resume appraisal procedure in bright any embodiment.
The multiple embodiments of the present invention are described in detail above, but the present invention is not restricted to these specific embodiment,
Those skilled in the art can make a variety of variants and modifications embodiments on the basis of present inventive concept, these modifications and repair
Changing should all fall within scope of the present invention.
Claims (8)
1. a kind of resume appraisal procedure, including：
Multiple first resumes corresponding with multiple first feedback data including the first feedback action item are obtained, to each first letter
Go through construction feature vector and be used as negative sample, the first feedback action item represents that first resume is associated with the first feedback and moves
Make；
Multiple second resumes corresponding with multiple second feedback data including the second feedback action item are obtained, to each second letter
Go through construction feature vector and be used as positive sample, the second feedback action item represents that second resume is associated with the second feedback and moves
Make, first feedback action is the feedback action of the previous stage of second feedback action；
Come Training valuation model learning described eigenvector and the second feedback action item using the positive sample and negative sample
Between mapping relations；And
Calculated by feature vector of the assessment models based on resume to be assessed according to the mapping relations described to be assessed
Resume is associated with the probability of the second feedback action in the case where being associated with the first feedback action.
2. the method as described in claim 1, further includes：
The 3rd feedback data of the resume to be assessed is obtained, whether the described second feedback is included according to the 3rd feedback data
The probability of action item and the resume to be assessed, corrects the mapping relations that the assessment models are learnt.
3. the method for claim 1, wherein each first resume or the second resume construction feature vector are included：
Made with first resume or the second resume with the correlation being associated between other resumes of second feedback action
One of it is characterized to build the feature vector of the first resume or the second resume.
4. the method for claim 1, wherein first feedback action or the second feedback action are resume selection person's pin
To first resume or the second resume actions taken.
5. the method for claim 1, wherein first feedback data and/or the second feedback data further include auxiliary
Information, the probabilistic correlation is in the auxiliary information.
6. method as claimed in claim 5, wherein, the auxiliary information include the first job information, the first trade information and/
Or feedback time information.
7. such as the method any one of claim 1-6, further include：
For the more parts of resumes to be assessed, the probability based on each resume to be assessed to each resume to be assessed into
Row sequence.
8. a kind of resume apparatus for evaluating, including processor, it is characterised in that the processor run predetermined computer instruction with
Perform the resume appraisal procedure as any one of claim 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711321692.XA CN108009735B (en) | 2017-12-12 | 2017-12-12 | Resume evaluation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711321692.XA CN108009735B (en) | 2017-12-12 | 2017-12-12 | Resume evaluation method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108009735A true CN108009735A (en) | 2018-05-08 |
| CN108009735B CN108009735B (en) | 2022-07-22 |
Family
ID=62058705
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711321692.XA Active CN108009735B (en) | 2017-12-12 | 2017-12-12 | Resume evaluation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108009735B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109034896A (en) * | 2018-07-23 | 2018-12-18 | 北京奇艺世纪科技有限公司 | Crowd's prediction technique and device are launched in a kind of advertisement |
| CN109345201A (en) * | 2018-09-20 | 2019-02-15 | 成都辰语轩科技有限公司 | Human Resources Management Method, device, electronic equipment and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120123956A1 (en) * | 2010-11-12 | 2012-05-17 | International Business Machines Corporation | Systems and methods for matching candidates with positions based on historical assignment data |
| CN105488662A (en) * | 2016-01-07 | 2016-04-13 | 北京歌利沃夫企业管理有限公司 | Bi-directional recommendation-based online recruitment system |
| CN107194762A (en) * | 2017-05-05 | 2017-09-22 | 腾讯科技（深圳）有限公司 | The vehicles recommend method, system and its equipment |
| CN107291715A (en) * | 2016-03-30 | 2017-10-24 | 阿里巴巴集团控股有限公司 | Resume appraisal procedure and device |
-
2017
- 2017-12-12 CN CN201711321692.XA patent/CN108009735B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120123956A1 (en) * | 2010-11-12 | 2012-05-17 | International Business Machines Corporation | Systems and methods for matching candidates with positions based on historical assignment data |
| CN105488662A (en) * | 2016-01-07 | 2016-04-13 | 北京歌利沃夫企业管理有限公司 | Bi-directional recommendation-based online recruitment system |
| CN107291715A (en) * | 2016-03-30 | 2017-10-24 | 阿里巴巴集团控股有限公司 | Resume appraisal procedure and device |
| CN107194762A (en) * | 2017-05-05 | 2017-09-22 | 腾讯科技（深圳）有限公司 | The vehicles recommend method, system and its equipment |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109034896A (en) * | 2018-07-23 | 2018-12-18 | 北京奇艺世纪科技有限公司 | Crowd's prediction technique and device are launched in a kind of advertisement |
| CN109345201A (en) * | 2018-09-20 | 2019-02-15 | 成都辰语轩科技有限公司 | Human Resources Management Method, device, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108009735B (en) | 2022-07-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107993019A (en) | A kind of resume appraisal procedure and device | |
| Anjomshoae et al. | An integrated AHP-based scheme for performance measurement in humanitarian supply chains | |
| Larson et al. | Project Management: The Managerial Process 6e | |
| George | The lean six sigma guide to doing more with less: cut costs, reduce waste, and lower your overhead | |
| Tseng et al. | Enhancing enterprise agility by deploying agile drivers, capabilities and providers | |
| CN110727875B (en) | Intelligent distribution method and system for legal case agent | |
| Phillips et al. | High-impact human capital strategy: Addressing the 12 major challenges today's organizations face | |
| Svanberg et al. | Beyond AI Exposure: Which Tasks are Cost-Effective to Automate with Computer Vision? | |
| US11797941B2 (en) | System and methodologies for candidate analysis utilizing psychometric data and benchmarking | |
| CN108182512A (en) | A kind of resume appraisal procedure and device | |
| CN108009735A (en) | A kind of resume appraisal procedure and device | |
| WO2021202407A1 (en) | Computer platform implementing many-to-many job marketplace | |
| CN108182513A (en) | A kind of resume appraisal procedure and device | |
| CN108153829A (en) | A kind of resume appraisal procedure and device | |
| WO2020053737A1 (en) | System and method for adapting an organization to future workforce requirements | |
| Kazemifard et al. | A Novel Hybrid Model for Technology Strategy Formulating in High-tech Industries under Uncertainty: A Case Study | |
| Zozulov et al. | Brand audit: types, approaches, methodologies | |
| KR102044219B1 (en) | service method for customized curriculum design process based on competency-based education | |
| PANNEERSELVAM et al. | Quality management | |
| Zhang et al. | An intelligent scheduling model of computer human resources in complex scenarios based on artificial intelligence | |
| Hadden | A customer profiling methodology for churn prediction | |
| KR20180025674A (en) | SW Market Information Service System | |
| Amre et al. | Decision Making Approach in Recruitment using Neuro-Fuzzy System | |
| KR102677614B1 (en) | Methods and systems for matching and rewarding human resources by analyzing individual competencies and mbti tendencies based on artificial intelligence | |
| Wickramasinghe et al. | Digital platform to empower the self-employment in Sri Lanka |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP01 | Change in the name or title of a patent holder | ||
| CP01 | Change in the name or title of a patent holder |
Address after: 100041 B-0035, 2 floor, 3 building, 30 Shixing street, Shijingshan District, Beijing. Patentee after: Douyin Vision Co.,Ltd. Address before: 100041 B-0035, 2 floor, 3 building, 30 Shixing street, Shijingshan District, Beijing. Patentee before: Tiktok vision (Beijing) Co.,Ltd. Address after: 100041 B-0035, 2 floor, 3 building, 30 Shixing street, Shijingshan District, Beijing. Patentee after: Tiktok vision (Beijing) Co.,Ltd. Address before: 100041 B-0035, 2 floor, 3 building, 30 Shixing street, Shijingshan District, Beijing. Patentee before: BEIJING BYTEDANCE NETWORK TECHNOLOGY Co.,Ltd. |