CN110263148A - Intelligent resume selection method and device - Google Patents
Intelligent resume selection method and device Download PDFInfo
- Publication number
- CN110263148A CN110263148A CN201910565804.9A CN201910565804A CN110263148A CN 110263148 A CN110263148 A CN 110263148A CN 201910565804 A CN201910565804 A CN 201910565804A CN 110263148 A CN110263148 A CN 110263148A
- Authority
- CN
- China
- Prior art keywords
- resume
- ability
- label
- text
- obtains
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/31—Indexing; Data structures therefor; Storage structures
- G06F16/316—Indexing structures
- G06F16/322—Trees
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/335—Filtering based on additional data, e.g. user or group profiles
Abstract
The present invention provides a kind of intelligent resume selection method and devices, intelligent resume selection device includes the corresponding ability label in Filtration Goal post from the ability tag library being pre-created, generative capacity tag tree, wherein, the ability tag library is the set of ability label corresponding to the corresponding various ability needs in each post；According to the ability tag tree from the resume text of acquisition extractability label；Resume selection is carried out according to the ability tag tree of the ability label and the target post that extract in the resume text.The application can carry out natural language analysis based on requirement capability tag library and resume nature text, delineate the ability portrait of resume, realize the intelligent Evaluation of resume.On the one hand, reduce the human input of talent's screening to the full extent, on the other hand, using the existing software and hardware resources of bank, further artificial intelligence is pushed to assist production decision.
Description
Technical field
This application involves artificial intelligence technologys, in particular to information sifting technology, and in particular to a kind of intelligence resume selection
Method and device.
Background technique
Each enterprises and institutions need labor intensive to go to check a large amount of post application resume, require simultaneously during recruitment
Reviewer has very high post cognitive ability.Existing resume selection relies on corresponding post Recruiting Specialist more, and naked eyes consult letter
It goes through, the ability agreed with post is found in big section text presentation, to achieve the purpose that judge resume superiority and inferiority.It is artificial to examine
The process judging basis for readding resume is more subjective, and different reviewers can have cognizance hierarchy, and applicant cannot be described by resume
Objective superiority and inferiority is obtained to compare.In addition, existing resume selection process data reusability is low, it is unable to satisfy multiple difference retrieval
The quick access of condition.
Summary of the invention
For the problems of the prior art, the application provides a kind of intelligent resume selection method and device, can be in resume
The ability label and competency degree of all applicant's resumes of rapidly extracting when screening are people to realize resume Fast Evaluation
Just screening provides decision-making foundation.
In order to solve the above technical problems, the application the following technical schemes are provided:
In a first aspect, the application provides a kind of intelligent resume selection method, comprising:
The corresponding ability label in Filtration Goal post from the ability tag library being pre-created, generative capacity tag tree,
In, the ability tag library is the set of ability label corresponding to the corresponding various ability needs in each post；
According to the ability tag tree from the resume text of acquisition extractability label；
Resume sieve is carried out according to the ability tag tree of the ability label and the target post that extract in the resume text
Choosing.
Further, the intelligent resume selection method, further includes: create the ability tag library.
Further, the creation ability tag library, comprising:
Obtain the corresponding ability need information in each post；
Ability demand information is subjected to domain classification, and each field is divided into multiple subclasses；
Assign the corresponding different abilities of each subclass to corresponding different ability label；
Determine that post corresponds to the weight of each ability label.
Further, the intelligent resume selection method, the creation ability tag library further include:
The meaning and scope of ability label based on imparting determine the ability label according to the describing mode of history resume
Matching rule；
The participle text of the matching rule is subjected to text vector by coding mode, obtains ability label vector.
Further, extractability label in the resume text from acquisition, comprising:
Word segmentation processing is carried out to the sentence of the resume text, obtains resume participle text；
Text is segmented to the resume, text vector is carried out by coding mode, obtains text vector；
Calculate the similarity of ability tag tree corresponding the ability label vector and text vector of the target post；
Ability label in resume text is extracted according to preset similarity threshold.
Further, the ability label according to the ability label extracted in the resume text and the target post
Tree carries out resume selection, comprising:
It sums respectively to corresponding each ability label weighted value of each subclass in the resume, obtains commenting for each subclass
Point；
The scoring of each subclass is summed, the overall score of the resume is obtained；
Resume selection is carried out according to the overall score of preset scoring threshold value and the resume, the scoring threshold value is according to
The ability tag tree of target post determines.
Further, the ability label according to the ability label extracted in the resume text and the target post
Tree carries out resume selection, comprising:
The ability label extracted in the resume text is corresponding according to the corresponding weight in the post and ability label
Competency degree is weighted, and obtains ability label weighted value, wherein the competency degree is the corresponding ability level of ability label；
It sums respectively to corresponding each ability label weighted value of each subclass in the resume, obtains commenting for each subclass
Point；
The scoring of each subclass is summed, the overall score of the resume is obtained；
Resume selection is carried out according to the overall score of preset scoring threshold value and the resume, the scoring threshold value is according to
The ability tag tree of target post determines.
Further, the intelligent resume selection method, further includes:
Resume text is split as sentence list；
Word segmentation processing is carried out to each sentence in the sentence list, obtains the word list of each sentence；
Part of speech analysis is carried out to the word list；
Syntactic analysis is carried out to the word list, obtains the semantic relation in same sentence between different words；
The matching word of the ability label of extraction is searched in the resume text, and obtains what the matching word occurred
Word order coordinate；
From the word order coordinate, left and right traversal is carried out, the matching word is found according to the part of speech, semantic relation
Adjunctival；
Numerical value intensity is converted by the adjunctival, obtains competency degree.
Second aspect, the application provide a kind of intelligent resume selection device, comprising:
Tag tree generation unit, for the corresponding ability mark in Filtration Goal post from the ability tag library being pre-created
Label, generative capacity tag tree, wherein the ability tag library is ability mark corresponding to the corresponding various ability needs in each post
The set of label；
Tag extraction unit, for according to the ability tag tree from the resume text of acquisition extractability label；
Resume selection unit, for the ability according to the ability label and the target post that are extracted in the resume text
Tag tree carries out resume selection.
Further, the intelligent resume selection device, further includes: tag library creating unit, for creating the energy
Power tag library.
Further, the tag library creating unit includes:
Data obtaining module, for obtaining the corresponding ability need information in each post；
Each field for ability demand information to be carried out domain classification, and is divided into multiple subclasses by information categorization module；
Label determining module, for assigning the corresponding different abilities of each subclass to corresponding different ability label；
Weight determining module, for determining that post corresponds to the weight of each ability label.
Further, the tag library creating unit, further includes:
Matching rule setting module, for the meaning and scope of the ability label based on imparting, according to retouching for history resume
The mode of stating determines the matching rule of the ability label；
First text vector module, for by the participle text of the matching rule by coding mode carry out text to
Quantization, obtains ability label vector.
Further, the tag extraction unit, comprising:
First text word segmentation module carries out word segmentation processing for the sentence to the resume text, obtains resume participle text
This；
Second text vector module carries out text vector by coding mode for segmenting text to the resume,
Obtain text vector；
Similarity assessment module, the corresponding ability label vector of ability tag tree for calculating the target post
With the similarity of text vector；
Label filtration module, for extracting ability label in resume text according to preset similarity threshold.
Further, the resume selection unit, comprising:
First subclass label grading module is weighted for corresponding each ability label to each subclass in the resume
Value is summed respectively, obtains the scoring of each subclass；
First general comment sub-module obtains the overall score of the resume for the scoring of each subclass to be summed；
First resume selection module, for carrying out resume sieve according to the overall score of preset scoring threshold value and the resume
Choosing, the scoring threshold value are determined according to the ability tag tree of the target post.
Further, the resume selection unit, comprising:
Label weighting block, the ability label for will extract in the resume text is according to the corresponding weight in the post
And the corresponding competency degree of ability label is weighted, and obtains ability label weighted value, wherein the competency degree is ability mark
Sign corresponding ability level；
Second subclass label grading module is weighted for corresponding each ability label to each subclass in the resume
Value is summed respectively, obtains the scoring of each subclass；
Second general comment sub-module obtains the overall score of the resume for the scoring of each subclass to be summed；
Second resume selection module, for carrying out resume sieve according to the overall score of preset scoring threshold value and the resume
Choosing, the scoring threshold value are determined according to the ability tag tree of the target post.
Further, the intelligent resume selection device, further includes:
Text splits module, for resume text to be split as sentence list；
Second text word segmentation module obtains each for carrying out word segmentation processing to each sentence in the sentence list
The word list of sentence；
Part-of-speech tagging module, for carrying out part of speech analysis to the word list；
Syntactic analysis module obtains the language in same sentence between different words for carrying out syntactic analysis to the word list
Adopted relationship；
Matching word positioning unit, the matching word of the ability label for searching for extraction in the resume text, and
Obtain the word order coordinate that the matching word occurs；
Adjunctival extraction unit, for carrying out left and right traversal from the word order coordinate, according to the part of speech, language
Adopted relationship finds the adjunctival of the matching word；
Adjunctival quantifying unit obtains competency degree for converting numerical value intensity for the adjunctival.
The third aspect, the application provides a kind of electronic equipment, including memory, processor and storage are on a memory and can
The computer program run on a processor, the processor realize the intelligent resume selection method when executing described program
The step of.
Fourth aspect, the application provide a kind of computer readable storage medium, are stored thereon with computer program, the calculating
The step of intelligent resume selection method is realized when machine program is executed by processor.
As shown from the above technical solution, the application provides a kind of intelligent resume selection method and device, intelligent resume selection
Method includes: the corresponding ability label in Filtration Goal post from the ability tag library being pre-created, generative capacity tag tree,
In, the ability tag library is the set of ability label corresponding to the corresponding various ability needs in each post；According to the energy
Power tag tree extractability label from the resume text of acquisition；According to the ability label extracted in the resume text with it is described
The ability tag tree of target post carries out resume selection, can be with the ability label and ability of all applicant's resumes of rapidly extracting
Intensity provides decision-making foundation to realize resume Fast Evaluation for talent's screening.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
It obtains other drawings based on these drawings.
Fig. 1 is the intelligent resume selection method flow diagram of the embodiment of the present application；
Fig. 2 is the flow chart of the creation ability tag library of the embodiment of the present application；
Fig. 3 is the flow chart of extractability label in the resume text from acquisition of the embodiment of the present application；
Fig. 4 is the energy according to the ability label and the target post that extract in the resume text of the embodiment of the present application
The flow chart of power tag tree progress resume selection；
Fig. 5 is the energy according to the ability label and the target post that extract in the resume text of the embodiment of the present application
The flow chart of power tag tree progress resume selection；
Fig. 6 is the additional step flow chart of the intelligent resume selection method of the embodiment of the present application；
Fig. 7 is the structural schematic diagram of the intelligent resume selection device of the embodiment of the present application；
Fig. 8 is the structural schematic diagram of the tag library creating unit 704 of the embodiment of the present application；
Fig. 9 is the additional structure schematic diagram of the tag library creating unit 704 of the embodiment of the present application；
Figure 10 is the structural schematic diagram of the tag extraction unit 702 of the embodiment of the present application；
Figure 11 is the structural schematic diagram of the resume selection unit 703 of the embodiment of the present application；
Figure 12 is the structural schematic diagram of the resume selection unit 703 of the embodiment of the present application；
Figure 13 is the additional structure schematic diagram of the intelligent resume selection device of the embodiment of the present application.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
More subjective for the process judging basis of existing manual reviews' resume, it is poor that different reviewers can have cognition
Different, applicant cannot be described to obtain objective superiority and inferiority by resume and be compared and the multiplexing of existing resume selection process data
The problem of rate is low, is unable to satisfy the quick access of multiple difference search condition, the application provide a kind of intelligent resume selection method
And device, resume sieve is carried out according to the matching degree of ability label needed for the ability label and target post that extract in resume text
Choosing.
Fig. 1 is the intelligent resume selection method flow diagram of the embodiment of the present application, as shown in Figure 1, the intelligence resume selection side
Method includes:
S101: the corresponding ability label in Filtration Goal post, generative capacity label from the ability tag library being pre-created
Tree, wherein the ability tag library is the set of ability label corresponding to the corresponding various ability needs in each post；
Multiple ability labels, such as program capability label, sociability label etc. are generally comprised in each target post, this
A little ability labels constitute the ability tag tree of the target post.
S102: according to the ability tag tree from the resume text of acquisition extractability label；
In order to screen more matched resume, the extractability label from the resume text of acquisition is needed, the ability mark
Label belong to the ability tag tree of target post.
S103: letter is carried out according to the ability tag tree of the ability label and the target post that extract in the resume text
Go through screening.
Process as shown in Figure 1 is it is found that the application Filtration Goal post pair first from the ability tag library being pre-created
The ability label answered, generative capacity tag tree；Then according to the ability tag tree from the resume text of acquisition extractability
Label；Resume sieve is finally carried out according to the ability tag tree of the ability label and the target post that extract in the resume text
Choosing.In this way, can be with the ability label and competency degree of all applicant's resumes of rapidly extracting, to realize that resume is fast
Speed evaluation provides decision-making foundation for talent's screening.
In one embodiment, intelligent resume selection method further include: the step of creating the ability tag library.
When it is implemented, as shown in Fig. 2, creating the ability tag library step and including:
S201: the corresponding ability need information in each post, such as education educational background, contest, project experiences, technical capability are obtained
Many-sided demand such as point；
S202: carrying out domain classification for ability demand information, such as is divided into science, practices, technical ability, intelligence and art, personality, honor
The fields such as reputation；And each field is divided into multiple subclasses, such as subclass in technical ability includes computer level, L proficiency etc..
S203: the corresponding different abilities of each subclass are assigned to corresponding different ability label；
S204: determine that post corresponds to the weight of each ability label according to each post capability demand power.
For example, foreign language aptitude belongs to the stronger ability of demand, and the weight of corresponding ability label is general for translating post
It is higher.
In one embodiment, intelligent resume selection method further include:
The meaning and scope of ability label based on imparting determine the ability label according to the describing mode of history resume
Matching rule；For example, the programming languages such as C++, Java can be used as the matching rule of the ability label for program capability label.
The participle text of the matching rule is subjected to text vector by coding mode, obtains ability label vector；
Text vector common method has onehot, bow, tf-idf, word2vec, doc2vec, Hash vector etc.；Preferably, may be used
Text vector is converted to the orderly word list that the coding mode using bow exports text normalization unit.
In one embodiment, intelligent resume selection method further include: the step of extractability label from the resume text of acquisition
Suddenly.
When it is implemented, as shown in figure 3, extractability labelling step includes: from the resume text of acquisition
S301: carrying out word segmentation processing to the sentence of the resume text, obtains resume participle text；Specifically, according to industry
Proper nouns dictionary in business knowledge architecture field is segmented as sentence of the customized dictionary to each field, so that obtaining one has
Sequence word list；Dictionary is deactivated it is also possible to construct according to professional knowledge, filters the word that task is unrelated in word list, most
Resume participle text is obtained eventually.
S302: text is segmented to the resume, text vector is carried out by coding mode, obtain text vector；It is specific real
Shi Shi needs to carry out text vector by coding mode identical with the participle text of the matching rule.
S303: the phase of ability tag tree corresponding the ability label vector and text vector of the target post is calculated
Like degree；Common method includes the cosine value calculated between the text vector of two sections of texts, is judged according to the size of cosine value similar
Degree, the cosine value between similarity and vector is corresponding, and cosine value numerical value is bigger, and to represent similarity bigger.Preferably, it can adopt
The cosine value between the text vector of two sections of texts is calculated with docsim method.
S304: ability label in resume text is extracted according to preset similarity threshold.Specifically, can for it is related
The similarity of ability label is higher than the resume text of a certain threshold value, judges the phase of its same ability label within the scope of full dose resume
Like degree ranking, before ranking 30% resume text is stamped into the ability label.
In addition, can also judge one by one current for the resume for still not stamping ability label after above-mentioned steps
The ability label similarity of resume text, if the similarity ranking of optimum capacity label stamps the ability preceding 60%, for it
Label realizes secondary selection.
In one embodiment, as shown in figure 4, according to the ability label and the target post extracted in the resume text
Ability tag tree carry out resume selection the step of include:
S401: it sums respectively to corresponding each ability label weighted value of each subclass in the resume, obtains each son
The scoring of class；
S402: the scoring of each subclass is summed, the overall score of the resume is obtained；
For above-mentioned overall score, the overall score of all resumes can be normalized to 0 to 100 by minimum point and best result
In fraction range space, the score of all resumes becomes the conventional fractional within 100, and the application is without being limited thereto.
S403: resume selection, the scoring threshold value root are carried out according to the overall score of preset scoring threshold value and the resume
It is determined according to the ability tag tree of the target post.
Specifically, the resume by overall score lower than scoring threshold value filters out, and filters out suitable resume.
In one embodiment, as shown in figure 5, according to the ability label and the target post extracted in the resume text
Ability tag tree carry out resume selection the step of include:
S501: by the ability label extracted in the resume text according to the corresponding weight in the post and ability label pair
The competency degree answered is weighted, and obtains ability label weighted value, wherein the competency degree is the corresponding ability of ability label
It is horizontal；
S502: it sums respectively to corresponding each ability label weighted value of each subclass in the resume, obtains each son
The scoring of class；
S503: the scoring of each subclass is summed, the overall score of the resume is obtained；
For above-mentioned overall score, the overall score of all resumes can be normalized to 0 to 100 by minimum point and best result
In fraction range space, the score of all resumes becomes the conventional fractional within 100, and the application is without being limited thereto.
S504: resume selection, the scoring threshold value root are carried out according to the overall score of preset scoring threshold value and the resume
It is determined according to the ability tag tree of the target post.
In one embodiment, as shown in fig. 6, the step of intelligent resume selection method further include:
S601: resume text is split as sentence list；Specifically, can by each field of every resume according to point
Sentence punctuate such as fullstop, exclamation mark, question mark etc., splits into sentence list；
S602: word segmentation processing is carried out to each sentence in the sentence list, obtains the word list of each sentence；It is logical
Cross the stop words dictionary being loaded into, filter out the stop words being not concerned with, such as " ", "Yes" etc.；Filter out the special word being not concerned with
The word, such as " we " etc. of property；
S603: part of speech analysis is carried out to the word list；
Specifically, a word the most suitable can be put on to each word in sentence according to the contextual information of sentence
Property；Preferably, part-of-speech tagging can be realized by Hidden Markov Model, this method can pass through the probability distribution of ambiguous category part of speech
Certain disambiguation is carried out to ambiguity of POS；
S604: carrying out syntactic analysis to the word list, obtains the semantic relation in same sentence between different words；
Specifically, the sentence structure analysis method of the syntactic structure of determining sentence can be divided into according to target difference and determines sentence
The dependency analysis method of dependence between son, it is preferred to use interdependent syntactic analysis method；Currently, interdependent syntactic analysis side
Method is with respect to sentence structure analysis method, and training mark is simple, flexible and accuracy rate is high.
S605: the matching word of the ability label of extraction is searched in the resume text, and obtains the matching word
The word order coordinate of appearance；Word order coordinate is the position of ability label in the text.
S606: from the word order coordinate, carrying out left and right traversal, finds described according to the part of speech, semantic relation
Adjunctival with word；
Specifically, left and right traversal is carried out repeatedly from coordinate points by the word order list of coordinates；Ergodic process
In, every to promote a word to corresponding direction, the part of speech for obtaining the word, syntactic analysis are as a result, judge this word by rule of combination
Whether the ornamental equivalent of capability goal word is belonged to.If certain direction searches non-modified exit in advance at participle and traverses when secondary.By certain
The immediate continuous ornamental equivalent in position on a direction is fused into the same adjunctival.
S607: numerical value intensity is converted by the adjunctival, obtains competency degree；
Specifically, adjunctival is converted in the way of presorting 0 to 5 numerical value intensity, if can not be identified as pre-
Trained classification is then identified as intensity 2；Merger is carried out according to same capabilities label, obtains the corresponding competency degree of ability label.
Based on the same inventive concept, the embodiment of the present application also provides a kind of intelligent resume selection device, it can be used for reality
Method described in existing above-described embodiment, as described in the following examples.The original solved the problems, such as due to intelligent resume selection device
Reason is similar to intelligent resume selection method, therefore the implementation of intelligent resume selection device may refer to based on intelligent resume selection side
The implementation of method, overlaps will not be repeated.Used below, predetermined function may be implemented in term " unit " or " module "
The combination of software and/or hardware.Although system described in following embodiment is preferably realized with software, hardware, or
The realization of the combination of person's software and hardware is also that may and be contemplated.
As shown in fig. 7, the application provides a kind of embodiment of intelligent resume selection device, comprising:
Tag tree generation unit 701, for the corresponding ability in Filtration Goal post from the ability tag library being pre-created
Label, generative capacity tag tree, wherein the ability tag library is ability corresponding to the corresponding various ability needs in each post
The set of label；
Tag extraction unit 702, for according to the ability tag tree from the resume text of acquisition extractability label；
Resume selection unit 703, for according to the ability label and the target post extracted in the resume text
Ability tag tree carries out resume selection.
Further, the intelligent resume selection device, further includes: tag library creating unit 704, it is described for creating
Ability tag library.
Further, as shown in figure 8, the tag library creating unit 704 includes:
Data obtaining module 7041, for obtaining the corresponding ability need information in each post；
Information categorization module 7042 for ability demand information to be carried out domain classification, and each field is divided into multiple
Subclass；
Label determining module 7043, for assigning the corresponding different abilities of each subclass to corresponding different ability label；
Weight determining module 7044, for determining that post corresponds to the weight of each ability label.
Further, as shown in figure 9, the tag library creating unit 704, further includes:
Matching rule setting module 7045, for the meaning and scope of the ability label based on imparting, according to history resume
Describing mode determine the matching rule of the ability label；
First text vector module 7046, for the participle text of the matching rule to be carried out text by coding mode
This vectorization obtains ability label vector.
Further, as shown in Figure 10, the tag extraction unit 702, comprising:
First text word segmentation module 7021 carries out word segmentation processing for the sentence to the resume text, standardizes original
The format of text；Specifically, according to proper nouns dictionary in professional knowledge building field, as customized dictionary to each field
Sentence participle, to obtain an orderly word list；It is constructed also according to professional knowledge and deactivates dictionary, filtered in word list
The unrelated word of task finally obtains resume participle text；
Second text vector module 7022, for carrying out text vector by coding mode to resume participle text
Change, obtains text vector；Text vector common method have onehot, bow, tf-idf, word2vec, doc2vec, Hash to
Quantization etc.；It is preferred that being converted to text vector using the orderly word list that the coding mode of bow exports text normalization unit.
Similarity assessment module 7023, the corresponding ability label of ability tag tree for calculating the target post
The similarity of vector and text vector；Common method is the cosine value calculated between the text vector of two sections of texts, and numerical value is bigger
It is bigger to represent similarity, it is preferred to use docsim method.
Label filtration module 7024, for extracting ability label in resume text according to preset similarity threshold.Specifically
, automated tag unit is directed to the resume text for being higher than a certain threshold value with the similarity of related capabilities label, judges it in full dose
Before ranking 30% resume text is stamped the ability label by the similarity ranking of same ability label within the scope of resume；For
The resume that ability label is not still stamped after above-mentioned steps judges that the ability label of current resume text is similar one by one
Degree, if the similarity ranking of optimum capacity label stamps the ability label preceding 60%, for it.
Further, as shown in figure 11, the resume selection unit 703, comprising:
First subclass label grading module 7031, for corresponding each ability label to each subclass in the resume
Weighted value is summed respectively, obtains the scoring of each subclass；Preferably, the ratio for calculating each subclass scoring, obtains the scoring of each subclass
Accounting.
First general comment sub-module 7032 obtains the overall score of the resume for the scoring of each subclass to be summed；It is excellent
Choosing, the first scoring of all resumes is normalized to by minimum point and best result in 0 to 100 fraction range space, Suo Youjian
The score gone through becomes the conventional fractional within 100.
First resume selection module 7033, for carrying out resume according to the overall score of preset scoring threshold value and the resume
Screening, the scoring threshold value are determined according to the ability tag tree of the target post.
Further, as shown in figure 12, the resume selection unit 703, comprising:
Label weighting block 7034, the ability label for will extract in the resume text are corresponding according to the post
Weight and the corresponding competency degree of ability label are weighted, and obtain ability label weighted value, wherein the competency degree is energy
The corresponding ability level of power label；
Second subclass label grading module 7035, for corresponding each ability label to each subclass in the resume
Weighted value is summed respectively, obtains the scoring of each subclass；Preferably, the ratio for calculating each subclass scoring, obtains the scoring of each subclass
Accounting.
Second general comment sub-module 7036 obtains the overall score of the resume for the scoring of each subclass to be summed；It is excellent
Choosing, the first scoring of all resumes is normalized to by minimum point and best result in 0 to 100 fraction range space, Suo Youjian
The score gone through becomes the conventional fractional within 100.
Second resume selection module 7037, for carrying out resume according to the overall score of preset scoring threshold value and the resume
Screening, the scoring threshold value are determined according to the ability tag tree of the target post.
Further, as shown in figure 13, the intelligent resume selection device further include:
Text split cells 705, for resume text to be split as sentence list；
Second text participle unit 706 obtains every for carrying out word segmentation processing to each sentence in the sentence list
The word list of a sentence；
Part-of-speech tagging unit 707, for carrying out part of speech analysis to the word list；It is given according to the contextual information of sentence
Each word in sentence puts on a part of speech the most suitable；Common part-of-speech tagging method is by Hidden Markov Model reality
Existing, this method can carry out certain disambiguation to ambiguity of POS by the probability distribution of ambiguous category part of speech.
Syntactic analysis unit 708 obtains in same sentence between different words for carrying out syntactic analysis to the word list
Semantic relation；According to target difference can be divided into the syntactic structure of determining sentence sentence structure analysis and determine sentence between according to
Deposit the dependency analysis of relationship；Currently, the sentence structure analysis research that tradition is made of grammer collection is gradually trained to mark
Replaced the dependency analysis research of simple and flexible high-accuracy；It is preferred that using interdependent syntactic analysis.
Matching word positioning unit 709, the matching word of the ability label for searching for extraction in the resume text,
And obtain the word order coordinate that the matching word occurs；
Adjunctival extraction unit 710, for carrying out left and right traversal from the word order coordinate, according to the part of speech,
Semantic relation finds the adjunctival of the matching word；Specifically, passing through the word order coordinate obtained in matching word positioning unit 755
List carries out left and right traversal repeatedly from coordinate points；It is every to promote a word to corresponding direction in traversal, obtain the word
As a result, judging whether this word belongs to the ornamental equivalent of capability goal word by rule of combination, certain direction is searched for part of speech, syntactic analysis
Rope is exited in advance at participle when time traversal to non-modified；Nearest continuous ornamental equivalent on some direction is fused into same
Adjunctival.
Adjunctival quantifying unit 711 obtains competency degree for converting numerical value intensity for the adjunctival.Tool
Body, adjunctival is converted in the way of presorting 0 to 5 numerical value intensity, if the classification of pre-training can not be identified as,
Then it is identified as intensity 2；Merger is carried out according to same capabilities label, obtains the corresponding competency degree of ability label.
It should be understood by those skilled in the art that, the embodiment of the present invention can provide as method, system or computer program
Product.Therefore, complete hardware embodiment, complete software embodiment or reality combining software and hardware aspects can be used in the present invention
Apply the form of example.Moreover, it wherein includes the computer of computer usable program code that the present invention, which can be used in one or more,
The computer program implemented in usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.) produces
The form of product.
The present invention be referring to according to the method for the embodiment of the present invention, the process of equipment (system) and computer program product
Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions
The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs
Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real
The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
Specific embodiment is applied in the present invention, and principle and implementation of the present invention are described, above embodiments
Explanation be merely used to help understand method and its core concept of the invention；At the same time, for those skilled in the art,
According to the thought of the present invention, there will be changes in the specific implementation manner and application range, in conclusion in this specification
Appearance should not be construed as limiting the invention.
Claims (18)
1. a kind of intelligence resume selection method characterized by comprising
The corresponding ability label in Filtration Goal post, generative capacity tag tree from the ability tag library being pre-created, wherein institute
State the set that ability tag library is ability label corresponding to the corresponding various ability needs in each post；
According to the ability tag tree from the resume text of acquisition extractability label；
Resume selection is carried out according to the ability tag tree of the ability label and the target post that extract in the resume text.
2. intelligence resume selection method according to claim 1, which is characterized in that further include: create the ability label
Library.
3. intelligence resume selection method according to claim 2, which is characterized in that the creation ability tag library packet
It includes:
Obtain the corresponding ability need information in each post；
Ability demand information is subjected to domain classification, and each field is divided into multiple subclasses；
Assign the corresponding different abilities of each subclass to corresponding different ability label；
Determine that post corresponds to the weight of each ability label.
4. intelligence resume selection method according to claim 3, which is characterized in that the creation ability tag library is also
Include:
The meaning and scope of ability label based on imparting, the matching of the ability label is determined according to the describing mode of history resume
Rule；
The participle text of the matching rule is subjected to text vector by coding mode, obtains ability label vector.
5. intelligence resume selection method according to claim 4, which is characterized in that mentioned in the resume text from acquisition
Take ability label, comprising:
Word segmentation processing is carried out to the sentence of the resume text, obtains resume participle text；
Text is segmented to the resume, text vector is carried out by coding mode, obtains text vector；
Calculate the similarity of ability tag tree corresponding the ability label vector and text vector of the target post；
Ability label in resume text is extracted according to preset similarity threshold.
6. intelligence resume selection method according to claim 3, which is characterized in that described to be mentioned according in the resume text
The ability label of the ability label and the target post that take carries out resume selection, comprising:
It sums respectively to corresponding each ability label weighted value of each subclass in the resume, obtains the scoring of each subclass；
The scoring of each subclass is summed, the overall score of the resume is obtained；
Resume selection is carried out according to the overall score of preset scoring threshold value and the resume, the scoring threshold value is according to the target
The ability tag tree in post determines.
7. intelligence resume selection method according to claim 4, which is characterized in that described to be mentioned according in the resume text
The ability label of the ability label and the target post that take carries out resume selection, comprising:
By the ability label extracted in the resume text according to the corresponding weight in the post and the corresponding ability of ability label
Intensity is weighted, and obtains ability label weighted value, wherein the competency degree is the corresponding ability level of ability label；
It sums respectively to corresponding each ability label weighted value of each subclass in the resume, obtains the scoring of each subclass；
The scoring of each subclass is summed, the overall score of the resume is obtained；
Resume selection is carried out according to the overall score of preset scoring threshold value and the resume, the scoring threshold value is according to the target
The ability tag tree in post determines.
8. intelligence resume selection method according to claim 7, which is characterized in that further include:
Resume text is split as sentence list；
Word segmentation processing is carried out to each sentence in the sentence list, obtains the word list of each sentence；
Part of speech analysis is carried out to the word list；
Syntactic analysis is carried out to the word list, obtains the semantic relation in same sentence between different words；
The matching word of the ability label of extraction is searched in the resume text, and obtains the word order that the matching word occurs
Coordinate；
From the word order coordinate, left and right traversal is carried out, the modification of the matching word is found according to the part of speech, semantic relation
Phrase；
Numerical value intensity is converted by the adjunctival, obtains competency degree.
9. a kind of intelligence resume selection device characterized by comprising
Tag tree generation unit, it is raw for the corresponding ability label in Filtration Goal post from the ability tag library being pre-created
At ability tag tree, wherein the ability tag library is ability label corresponding to the corresponding various ability needs in each post
Set；
Tag extraction unit, for according to the ability tag tree from the resume text of acquisition extractability label；
Resume selection unit, for the ability label according to the ability label and the target post that are extracted in the resume text
Tree carries out resume selection.
10. intelligence resume selection device according to claim 9, which is characterized in that further include: tag library creating unit,
For creating the ability tag library.
11. intelligence resume selection device according to claim 10, which is characterized in that the tag library creating unit packet
It includes:
Data obtaining module, for obtaining the corresponding ability need information in each post；
Each field for ability demand information to be carried out domain classification, and is divided into multiple subclasses by information categorization module；
Label determining module, for assigning the corresponding different abilities of each subclass to corresponding different ability label；
Weight determining module, for determining that post corresponds to the weight of each ability label.
12. intelligence resume selection device according to claim 11, which is characterized in that the tag library creating unit is also wrapped
It includes:
Matching rule setting module, for the meaning and scope of the ability label based on imparting, according to the description side of history resume
Formula determines the matching rule of the ability label；
First text vector module, for the participle text of the matching rule to be carried out text vector by coding mode
Change, obtains ability label vector.
13. intelligence resume selection device according to claim 12, which is characterized in that the tag extraction unit, comprising:
First text word segmentation module carries out word segmentation processing for the sentence to the resume text, obtains resume participle text；
Second text vector module carries out text vector by coding mode for segmenting text to the resume, obtains
Text vector；
Similarity assessment module, the corresponding ability label vector of ability tag tree and text for calculating the target post
The similarity of this vector；
Label filtration module, for extracting ability label in resume text according to preset similarity threshold.
14. intelligence resume selection device according to claim 11, which is characterized in that the resume selection unit, comprising:
First subclass label grading module, for corresponding each ability label weighted value to each subclass in the resume point
It does not sum, obtains the scoring of each subclass；
First general comment sub-module obtains the overall score of the resume for the scoring of each subclass to be summed；
First resume selection module, for carrying out resume selection, institute according to the overall score of preset scoring threshold value and the resume
Commentary divides threshold value to be determined according to the ability tag tree of the target post.
15. intelligence resume selection device according to claim 12, which is characterized in that the resume selection unit, comprising:
Label weighting block, the ability label for will extract in the resume text is according to the corresponding weight in the post and energy
The corresponding competency degree of power label is weighted, and obtains ability label weighted value, wherein the competency degree is ability label pair
The ability level answered；
Second subclass label grading module, for corresponding each ability label weighted value to each subclass in the resume point
It does not sum, obtains the scoring of each subclass；
Second general comment sub-module obtains the overall score of the resume for the scoring of each subclass to be summed；
Second resume selection module, for carrying out resume selection, institute according to the overall score of preset scoring threshold value and the resume
Commentary divides threshold value to be determined according to the ability tag tree of the target post.
16. intelligence resume selection device according to claim 15, which is characterized in that further include:
Text split cells, for resume text to be split as sentence list；
Second text participle unit obtains each sentence for carrying out word segmentation processing to each sentence in the sentence list
Word list；
Part-of-speech tagging unit, for carrying out part of speech analysis to the word list；
Syntactic analysis unit obtains the semantic pass in same sentence between different words for carrying out syntactic analysis to the word list
System；
Matching word positioning unit, the matching word of the ability label for searching for extraction in the resume text, and obtain
The word order coordinate that the matching word occurs；
Adjunctival extraction unit, for carrying out left and right traversal from the word order coordinate, according to the part of speech, semantic pass
System finds the adjunctival of the matching word；
Adjunctival quantifying unit obtains competency degree for converting numerical value intensity for the adjunctival.
17. a kind of electronic equipment including memory, processor and stores the calculating that can be run on a memory and on a processor
Machine program, which is characterized in that the processor realizes claim 1 to 8 described in any item intelligence letters when executing described program
The step of going through screening technique.
18. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program quilt
The step of claim 1 to 8 described in any item intelligent resume selection methods are realized when processor executes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910565804.9A CN110263148A (en) | 2019-06-27 | 2019-06-27 | Intelligent resume selection method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910565804.9A CN110263148A (en) | 2019-06-27 | 2019-06-27 | Intelligent resume selection method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110263148A true CN110263148A (en) | 2019-09-20 |
Family
ID=67922066
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910565804.9A Pending CN110263148A (en) | 2019-06-27 | 2019-06-27 | Intelligent resume selection method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110263148A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111507685A (en) * | 2020-04-14 | 2020-08-07 | 上海学为信息科技有限公司 | Method for adjusting matching conditions and promoting post matching based on user behaviors |
| CN111680895A (en) * | 2020-05-26 | 2020-09-18 | 中国平安财产保险股份有限公司 | Data automatic labeling method and device, computer equipment and storage medium |
| WO2021169111A1 (en) * | 2020-02-28 | 2021-09-02 | 平安国际智慧城市科技股份有限公司 | Resume screening method and apparatus, computer device and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130254129A1 (en) * | 2007-10-05 | 2013-09-26 | Martin Perlmutter | Technological solution to interview inefficiency |
| CN103544312A (en) * | 2013-11-04 | 2014-01-29 | 成都数之联科技有限公司 | Employment information matching method based on social network |
| CN107590133A (en) * | 2017-10-24 | 2018-01-16 | 武汉理工大学 | The method and system that position vacant based on semanteme matches with job seeker resume |
| CN109165295A (en) * | 2018-09-27 | 2019-01-08 | 天涯社区网络科技股份有限公司 | A kind of intelligence resume appraisal procedure |
| CN109213999A (en) * | 2018-08-20 | 2019-01-15 | 成都佳发安泰教育科技股份有限公司 | A kind of subjective item methods of marking |
| CN109582704A (en) * | 2018-10-17 | 2019-04-05 | 龙马智芯（珠海横琴）科技有限公司 | Recruitment information and the matched method of job seeker resume |
-
2019
- 2019-06-27 CN CN201910565804.9A patent/CN110263148A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130254129A1 (en) * | 2007-10-05 | 2013-09-26 | Martin Perlmutter | Technological solution to interview inefficiency |
| CN103544312A (en) * | 2013-11-04 | 2014-01-29 | 成都数之联科技有限公司 | Employment information matching method based on social network |
| CN107590133A (en) * | 2017-10-24 | 2018-01-16 | 武汉理工大学 | The method and system that position vacant based on semanteme matches with job seeker resume |
| CN109213999A (en) * | 2018-08-20 | 2019-01-15 | 成都佳发安泰教育科技股份有限公司 | A kind of subjective item methods of marking |
| CN109165295A (en) * | 2018-09-27 | 2019-01-08 | 天涯社区网络科技股份有限公司 | A kind of intelligence resume appraisal procedure |
| CN109582704A (en) * | 2018-10-17 | 2019-04-05 | 龙马智芯（珠海横琴）科技有限公司 | Recruitment information and the matched method of job seeker resume |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021169111A1 (en) * | 2020-02-28 | 2021-09-02 | 平安国际智慧城市科技股份有限公司 | Resume screening method and apparatus, computer device and storage medium |
| CN111507685A (en) * | 2020-04-14 | 2020-08-07 | 上海学为信息科技有限公司 | Method for adjusting matching conditions and promoting post matching based on user behaviors |
| CN111680895A (en) * | 2020-05-26 | 2020-09-18 | 中国平安财产保险股份有限公司 | Data automatic labeling method and device, computer equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107491531B (en) | Chinese network comment sensibility classification method based on integrated study frame | |
| Meredith | Computational music analysis | |
| Jamal et al. | Poetry classification using support vector machines | |
| Valakunde et al. | Multi-aspect and multi-class based document sentiment analysis of educational data catering accreditation process | |
| Nejat et al. | Exploring joint neural model for sentence level discourse parsing and sentiment analysis | |
| CN110263148A (en) | Intelligent resume selection method and device | |
| CN110175246A (en) | A method of extracting notional word from video caption | |
| CN115357719B (en) | Power audit text classification method and device based on improved BERT model | |
| Zhang et al. | Automated classification of computing education questions using Bloom’s taxonomy | |
| CN103885933A (en) | Method and equipment for evaluating text sentiment | |
| Nassiri et al. | Arabic readability assessment for foreign language learners | |
| Stemle et al. | Using language learner data for metaphor detection | |
| Szarvas et al. | Learning to rank lexical substitutions | |
| Smitha et al. | Meme classification using textual and visual features | |
| CN108090099A (en) | A kind of text handling method and device | |
| Jishan et al. | Natural language description of images using hybrid recurrent neural network | |
| CN114912425A (en) | Presentation generation method and device | |
| CN111897955B (en) | Comment generation method, device, equipment and storage medium based on encoding and decoding | |
| Balouchzahi et al. | LA-SACo: A study of learning approaches for sentiments analysis inCode-mixing texts | |
| Pathuri et al. | Feature based sentimental analysis for prediction of mobile reviews using hybrid bag-boost algorithm | |
| JP5032600B2 (en) | Document readability evaluation program and document readability evaluation apparatus | |
| Tikhonova et al. | NLP methods for automatic candidate’s cv segmentation | |
| CN112905796B (en) | Text emotion classification method and system based on re-attention mechanism | |
| Tekumalla et al. | Leveraging Large Language Models and Weak Supervision for Social Media data annotation: an evaluation using COVID-19 self-reported vaccination tweets | |
| Bekmanova et al. | Sentiment analysis model based on the word structural representation |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190920 |