CN109448792A - Choosing method, device and the electronic equipment of interpreter's gene - Google Patents
Choosing method, device and the electronic equipment of interpreter's gene Download PDFInfo
- Publication number
- CN109448792A CN109448792A CN201811096578.6A CN201811096578A CN109448792A CN 109448792 A CN109448792 A CN 109448792A CN 201811096578 A CN201811096578 A CN 201811096578A CN 109448792 A CN109448792 A CN 109448792A
- Authority
- CN
- China
- Prior art keywords
- interpreter
- genome
- gene
- maximum
- value
- 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
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)
Abstract
The embodiment of the present invention provides choosing method, device and the electronic equipment of a kind of interpreter's gene, this method comprises: choosing multiple interpreter's genomes respectively；For each interpreter's genome, multiple successful match rate samples are obtained, and calculate the mean value and standard deviation of the corresponding successful match rate of interpreter's genome accordingly；The corresponding contribution genome of the maximum in all mean values is chosen as maximum contribution genome, and its mean value and standard deviation are respectively defined as Largest Mean and maximum standard deviation；Its corresponding Z value is calculated based on its corresponding mean value and standard deviation and Largest Mean and maximum standard deviation for each contribution genome in addition to maximum contribution genome；Based on the corresponding Z value of each interpreter's genome, the gene met in the gene in the interpreter's genome to impose a condition and maximum interpreter's genome is merged, the interpreter's gene finally chosen is obtained.The embodiment of the present invention enables to the interpreter's gene selected that can preferably embody the otherness between interpreter.
Description
Technical field
The present embodiments relate to technical field of data processing, more particularly, to a kind of interpreter's gene choosing method,
Device and electronic equipment.
Background technique
Information age and networking have changed a lot translation mode.Platform is managed using translation flow,
Talent's data is stored according to different objects, to match most suitable interpreter according to contribution to be translated.Different interpreters, institute
The key message for including is not quite similar, then according to these key messages, can match most suitable translation contribution for interpreter, thus
Effectively improve translation efficiency and translation accuracy.
Interpreter matches with the gene of contribution to be referred to contribution gene and interpreter's gene under set strategy through Matching Model,
It is embodied as the process that contribution finds best interpreter.Selected is used to carry out the matched interpreter's gene of gene and other interpreter's genes
It compares, it should can preferably embody the otherness of interpreter, could be so that contribution to be translated is matched to the interpreter being more suitable for.
Interpreter's gene is referred mainly to by carrying out analytical calculation, quantification treatment, accessed presence to interpreter's characteristic attribute
It is combined in key message specific interpreter, that be different from other interpreters, unique.There are many sources of interpreter's gene,
In the social epoch, all data of the every act and every move of interpreter can extract gene.
Interpreter's gene is present in all interpreters of management platform, and different interpreters have different interpreter's genes.Due to tool
Body application difference, presently, there are interpreter/manuscript gene matching algorithm selection interpreter gene to be matched carry out matching meter
When calculation, the corresponding assortment of genes is often rule of thumb selected.
But in interpreter's course of work, gene can with the promotion of ability, the increase of time, accumulation of knowledge and send out
Raw corresponding variation.I.e. with the processing of task, examine and revise evaluation with QC, accumulation, the community activity of history corpus participation with
And the activities such as test of interpreter's ability, interpreter's gene will be constantly updated.Therefore, above-mentioned interpreter's gene selects mode empirically
It can have some limitations, cause the interpreter's gene selected that cannot embody the otherness between interpreter well.
Summary of the invention
In order to overcome the above problem or at least be partially solved the above problem, the embodiment of the present invention provides a kind of interpreter's base
Choosing method, device and the electronic equipment of cause, with so that the interpreter's gene selected can preferably embody the difference between interpreter
It is anisotropic.
In a first aspect, the embodiment of the present invention provides a kind of choosing method of interpreter's gene, comprising: arranged from alternative interpreter's gene
In table, the different gene of multiple groups is chosen respectively, constitutes multiple interpreter's genomes；For interpreter's genome described in each, carry out
Multiple matching result sampling, obtains multiple successful match rate samples, and be based on the multiple successful match rate sample, calculates this and translate
The mean value and standard deviation of the corresponding successful match rate of member's genome；Choose the corresponding interpreter's base of the maximum in all mean values
Because of group, it is defined as maximum interpreter's genome, and the mean value of maximum interpreter's genome is defined as Largest Mean, by institute
The standard deviation for stating maximum interpreter's genome is defined as maximum standard deviation；For in all interpreter's genomes except it is described most
Each described interpreter's genome except big interpreter's genome, is based on the corresponding mean value of interpreter's genome and the mark
The quasi- poor and Largest Mean and the maximum standard deviation, calculate the corresponding Z value of interpreter's genome；Based on all described
The corresponding Z value of each described interpreter's genome in interpreter's genome in addition to maximum interpreter's genome, from institute
There are the interpreter's genome for choosing in interpreter's genome and meeting and imposing a condition, and interpreter's gene that the satisfaction is imposed a condition
The gene in gene and maximum interpreter's genome in group merges, and obtains the interpreter's gene finally chosen；Wherein, the Z
Value indicates Z value in the verifying of large sample otherness.
Second aspect, the embodiment of the present invention provide a kind of selecting device of interpreter's gene, comprising: initial gene chooses mould
Block, for choosing the different gene of multiple groups respectively, constituting multiple interpreter's genomes from alternative interpreter's list of genes；First meter
Module is calculated, for multiple matching result sampling being carried out, obtaining multiple successful match rate samples for interpreter's genome described in each
This, and it is based on the multiple successful match rate sample, calculate the mean value and standard of the corresponding successful match rate of interpreter's genome
Difference；Maximum genome is chosen module and is defined as most for choosing the corresponding interpreter's genome of the maximum in all mean values
Big interpreter's genome, and the mean value of maximum interpreter's genome is defined as Largest Mean, by maximum interpreter's base
Because the standard deviation of group is defined as maximum standard deviation；Second computing module, for for being removed in all interpreter's genomes
Each described interpreter's genome except the maximum interpreter genome, based on the corresponding mean value of interpreter's genome and
The standard deviation and the Largest Mean and the maximum standard deviation calculate the corresponding Z value of interpreter's genome；Final base
Because choosing module, for based on being translated described in each in all interpreter's genomes in addition to maximum interpreter's genome
The corresponding Z value of member's genome chooses the interpreter's genome for meeting and imposing a condition from all interpreter's genomes, and will
Gene in the gene met in the interpreter's genome to impose a condition and maximum interpreter's genome merges, and obtains final
Interpreter's gene of selection；Wherein, the Z value indicates Z value in the verifying of large sample otherness.
The third aspect, the embodiment of the present invention provide a kind of electronic equipment, comprising: at least one processor, at least one
Manage device, communication interface and bus；The memory, the processor and the communication interface are completed mutual by the bus
Communication, the communication interface between the electronic equipment and interpreter's information equipment information transmission；In the memory
It is stored with the computer program that can be run on the processor, when the processor executes the computer program, is realized such as
The choosing method of interpreter's gene described in upper first aspect.
Fourth aspect, the embodiment of the present invention provide a kind of non-transient computer readable storage medium, the non-transient calculating
Machine readable storage medium storing program for executing stores computer instruction, and the computer instruction executes the computer described in first aspect as above
The choosing method of interpreter's gene.
Choosing method, device and the electronic equipment of interpreter's gene provided in an embodiment of the present invention, by being translated in advance from all
Multiple groups interpreter genome is chosen in interpreter's gene pool of member, and by calculating Z value corresponding to these interpreter's genomes, to choose
Z value meets the interpreter's genome to impose a condition, to choose as final as a result, enabling the interpreter's gene selected more preferable
Embody interpreter between otherness.
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 this hair
Bright some embodiments for those of ordinary skill in the art without creative efforts, can be with root
Other attached drawings are obtained according to these attached drawings.
Fig. 1 is the flow diagram of the choosing method of interpreter's gene provided in an embodiment of the present invention；
Fig. 2 is the pass according to interpreter's feature and interpreter's gene in the choosing method of interpreter's gene provided in an embodiment of the present invention
It is schematic diagram；
Fig. 3 is the structural schematic diagram of the selecting device of interpreter's gene provided in an embodiment of the present invention；
Fig. 4 is the entity structure schematic diagram of electronic equipment provided in an embodiment of the present invention.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiment of the embodiment of the present invention, instead of all the embodiments.Based on the embodiment in the embodiment of the present invention, ability
Domain those of ordinary skill every other embodiment obtained without making creative work, belongs to the present invention
The range of embodiment protection.
There are many sources of interpreter's gene, and in the social epoch, all data of the every act and every move of interpreter can be extracted
Gene out.Due to the difference of concrete application, presently, there are interpreter/manuscript gene matching algorithm in the to be matched of selection interpreter
When gene carries out matching primitives, the corresponding assortment of genes is often rule of thumb selected.But conventional method has certain limitation
Property, cause the interpreter's gene selected that cannot embody the otherness of interpreter well.
In view of the above-mentioned problems, the embodiment of the present invention from interpreter's gene pool of interpreter by choosing multiple groups interpreter gene in advance
Group, and by calculating Z value corresponding to these interpreter's genomes, meet the interpreter's genome to impose a condition to choose Z value, to make
It chooses for final as a result, the interpreter's gene selected is enabled preferably to embody the otherness between interpreter.Wherein, Z value table
Show Z value in the verifying of large sample otherness.
As the one aspect of the embodiment of the present invention, the present embodiment provides a kind of choosing methods of interpreter's gene, with reference to figure
1, it is the flow diagram of the choosing method of interpreter's gene provided in an embodiment of the present invention, comprising:
S101 chooses the different gene of multiple groups respectively, constitutes multiple interpreter's genomes from alternative interpreter's list of genes.
It is to be understood that before the interpreter's gene for carrying out the present embodiment is chosen, it in advance can be according to all properties of interpreter
Information establishes alternative interpreter's list of genes, may include in alternative interpreter's list of genes relevant to interpreter's particular community
All genes.Specifically, alternative interpreter's list of genes may be considered a gene pool, with gene for singly in the gene pool
Position storage has the gene relevant to interpreter's information extracted from all interpreters, i.e. interpreter's gene.Interpreter's gene refers mainly to pass through
Analytical calculation, quantification treatment carried out to interpreter's characteristic attribute, it is accessed be present in specific interpreter, be different from other interpreters
, the combination of unique key message.
According to alternative interpreter's list of genes in this step, multiple groups interpreter gene is chosen respectively, and translate respectively with each group
Member's gene constitutes a genome, and as interpreter's genome, which is the interpreter's genome selected.It is understood that
, can be from multiple interpreters in alternative interpreter's list of genes in random selection table when carrying out the selection of each group interpreter gene
Gene, then the interpreter's gene randomly selected using these may be constructed a genome, as interpreter's genome.
It is of course also possible to which predefined decimation rule, e.g., while extracting or successively extracting, interlacing is extracted or specified line number
It extracts, according to different interpreter's information extractions of gene characterization, quantity of extraction, etc..Carrying out actual extraction process later
When, the extraction for each group of interpreter's gene extracts phase from alternative interpreter's list of genes according to the decimation rule predetermined
The multiple genes answered.
For example, randomly selecting 3-5 different genes from alternative interpreter's list of genes, as one group of gene, one is constituted
A interpreter's genome.It then adopts in a like fashion, can choose or successively choose simultaneously multiple groups gene respectively, constitute multiple
Interpreter's genome, the embodiment of the present invention to this with no restriction.
S102 carries out multiple matching result sampling for each interpreter's genome, obtains multiple successful match rate samples
This, and multiple successful match rate samples are based on, calculate the mean value and standard deviation of the corresponding successful match rate of interpreter's genome.
It is to be understood that for the interpreter's genome selected for each group, it is thus necessary to determine that the matching of itself and contribution is imitated
Fruit, so that selection is more suitable for the matched interpreter's gene of gene.Meanwhile in order to without loss of generality, for each group of interpreter's genome,
Interpreter's genome can be inputted to given Matching Model, carry out multiple matching result sampling using given Matching Model, every time
Sampling can obtain a successful match rate sample.
It is understood that carrying out successful match rate sample using Matching Model for each group of interpreter's genome
When acquisition, the gene in this group of interpreter's genome is input in Matching Model, which can be according to the original text itself provided
Part gene calculates the successful match rate score of gene in interpreter's genome and contribution gene automatically and exports, then matches mould
The successful match rate score of type output can be used as a successful match rate sample.For same interpreter's genome, carry out more
Secondary above-mentioned matching result sampling process, then available multiple successful match rate samples.
Later, it for each interpreter selected genome, is obtained according to above-mentioned multiple matching result sampling
Multiple successful match rate samples calculate the comprehensive matching success rate of interpreter's genome, that is, calculate separately interpreter's genome pair
The mean value and standard deviation for the successful match rate answered.It is understood that each successful match rate sample, actually primary
The successful match rate score sampled with result.
For example, it is assumed that carrying out matching result sampling according to some interpreter's genome, n successful match rate sample difference is obtained
For p1,p2,...pn.The mean value of the corresponding successful match rate of interpreter's genome is then calculated according to it are as follows:
In formula, E (p) indicates the mean value of the corresponding successful match rate of interpreter's genome, piIndicate i-th of interpreter's genome
Successful match rate sample, n indicate the total number of the successful match rate sample for interpreter's genome acquisition.
On this basis, the standard deviation for calculating the corresponding successful match rate of interpreter's genome is as follows:
In formula, S indicates the standard deviation of the corresponding successful match rate of interpreter's genome, and E (p) indicates that interpreter's genome is corresponding
The mean value of successful match rate, piIndicate that i-th of successful match rate sample of interpreter's genome, n indicate to be directed to interpreter's genome
The total number of the successful match rate sample of acquisition.
Wherein, in one embodiment, multiple matching result sampling is being carried out, is obtaining the step of multiple successful match rate samples
Before rapid, further includes: according to the demand of the gene matching primitives precision with contribution to be translated, be set for matching result sampling
Total degree threshold value, i.e. given threshold.Then accordingly in actual samples, the number of acquisition successful match rate sample is total not less than this
Frequency threshold value.For example, for each interpreter's genome, it is desirable that the number of the successful match rate sample of extraction is no less than 50, then
The data 50 are preset total degree threshold value.
S103 chooses the corresponding interpreter's genome of the maximum in all mean values, is defined as maximum interpreter's genome, and will
The mean value of maximum interpreter's genome is defined as Largest Mean, and the standard deviation of maximum interpreter's genome is defined as maximum standard deviation.
It can be understood as, it is assumed that total group of number of interpreter's genome of selection is that m group can calculate then according to above-mentioned steps
To the corresponding mean value of m group and standard deviation.The embodiment of the present invention chooses value the maximum first from m mean value, and by the maximum
Interpreter's genome corresponding to person is defined as maximum interpreter's genome.Then correspondingly, the mean value of maximum interpreter's genome is determined
Justice is Largest Mean, with variable EmaxIt indicates, the standard deviation of maximum interpreter's genome is defined as maximum standard deviation, uses variable
SmaxIt indicates.
S104 is based on each interpreter's genome in all interpreter's genomes in addition to maximum interpreter's genome
It is corresponding to calculate interpreter's genome for the corresponding mean value of interpreter's genome and standard deviation and Largest Mean and maximum standard deviation
Z value；Wherein, Z value indicates Z value in the verifying of large sample otherness.
It is to be understood that in all interpreter's genomes are calculated according to above-mentioned steps in addition to maximum interpreter's genome
Each of on the basis of the mean value and standard deviation of the corresponding successful match rate of interpreter's genome that select, these are selected
Interpreter's genome, calculate its Z value.Specifically, for each of these interpreter's genomes interpreter's genome, according to it
The standard deviation and mean value of corresponding successful match rate, in conjunction with the corresponding Largest Mean E of maximum interpreter's genomemaxIt is marked with maximum
Quasi- difference Smax, calculate separately the corresponding Z value of interpreter's genome.
It is understood that the concept of Z value therein is the verifying of large sample otherness, the i.e. concept of Z value in Z verifying.Z
Inspection is the method for being generally used for large sample (i.e. sample size is greater than 30) mean difference and examining.It is with standard normal point
The theory of cloth come infer difference occur probability, so that whether the difference for comparing two average significant.When known standard deviation,
Whether the mean value for verifying one group of number is equal with a certain desired value.Translating of selecting is measured in the embodiment of the present invention using Z verifying
The matching difference verifying of member's genome, therefore the calculating of Z value is carried out to interpreter's genome that each is selected.
S105, it is corresponding based on each interpreter's genome in all interpreter's genomes in addition to maximum interpreter's genome
Z value chooses the interpreter's genome for meeting and imposing a condition, and interpreter's base that the satisfaction is imposed a condition from all interpreter's genomes
Because the gene in the gene and maximum interpreter's genome in group merges, the interpreter's gene finally chosen is obtained.
It is to be understood that the Z except the extragenic each interpreter's genome of maximum interpreter can be calculated according to above-mentioned steps
Value, may determine that otherness performance of each corresponding interpreter's genome when carrying out gene matching according to the Z value.Therefore, according to every
The corresponding Z value of a interpreter's genome, can use preset setting condition, judges that the corresponding interpreter's genome of the Z value is
The no otherness requirement for meeting setting.If conditions are not met, then reject it from each interpreter's genome selected, it is final remaining
All interpreter's genomes not being removed are satisfactory interpreter's genome, include that Z value is full in these interpreter's genomes
The interpreter's genome and maximum interpreter's genome that foot setting otherness requires.Gene in remaining all interpreter's genomes is taken
Out, and after removing the duplicate factor in these genes, one group of new gene is formed, i.e., as the interpreter's gene finally chosen.
For example, it is assumed that acquiring n successful match rate sample, these successful match rates in total for some interpreter's genome
Sample meets normal distribution.Meanwhile it presetting and having selected the setting condition of interpreter's gene for the confidence level for the gene selected is not
Lower than 95%, the Z value which corresponds to interpreter's genome is 1.96.Then, for each the interpreter's genome selected,
Its corresponding Z value is compared with 1.96, if Z value is greater than 1.96, the corresponding interpreter's genome of the Z value is rejected, otherwise,
Retain the corresponding interpreter's genome of the Z value.
Assuming that eliminating p according to above-mentioned treatment process from a interpreter's genome selected of all n and being unsatisfactory for setting
Interpreter's genome of condition, remaining n-p interpreter genome are to meet to impose a condition.Then, in this n-p interpreter's genome
In, may there are two or more than two interpreter's genomes in contain some interpreter's gene simultaneously.Therefore this n-p are translated
Whole interpreter's genes in member's genome take out, and are put into a new gene pool, multiple for occurring in the gene pool
Each interpreter's gene, rejects extra and only retains interpreter's gene.It is included in this final new gene pool
Multiple non-repetitive interpreter's genes, using these genes as the interpreter's gene finally chosen.
The choosing method of interpreter's gene provided in an embodiment of the present invention, by advance from interpreter's gene pool of all interpreters
Multiple groups interpreter genome is chosen, and by calculating Z value corresponding to these interpreter's genomes, meets setting condition to choose Z value
Interpreter's genome, as between final choose as a result, the interpreter's gene selected is enabled preferably to embody interpreter
Otherness.In addition, the interpreter chosen accordingly and contribution to be translated can be made to carry out more reasonable in gene matching application
Match, to effectively improve translation efficiency and translation accuracy rate.
Wherein, in one embodiment, from alternative interpreter's list of genes, the step of the different gene of multiple groups is chosen respectively
Before rapid, the method for the embodiment of the present invention further include:
Corresponding gene is extracted from the basic information of all interpreters, ability information, credit information and posterior infromation respectively,
And it is correspondingly formed basic information gene, ability information gene, credit information gene and the posterior infromation gene of interpreter；
Based on basic information gene, ability information gene, credit information gene and posterior infromation gene, alternative interpreter is constituted
List of genes.
It is to be understood that there are many sources of interpreter's gene, and in the social epoch, all data of the every act and every move of interpreter
Gene can be extracted, by the sources of interpreter's gene, the present embodiment extracts interpreter's gene from the following aspects,
Constitute alternative interpreter's list of genes:
Basic information, the personal relevant information of interpreter, such as name, age, location and contact information；
Ability information, the translation ability information that interpreter possesses, languages direction, industry field and the translation speed being such as good at
Deng；
Credit information, the credit information that interpreter accumulates during undertaking translations such as hand over original text rate and midway in time
Rate of sending back the manuscript etc.；
Posterior infromation, the correlation experience that interpreter accumulates during long campaigns translation, such as translate total number of word and
Total amount etc..
Above- mentioned information based on interpreter extract the corresponding corresponding gene of interpreter respectively, and according to above-mentioned various aspects, formation pair
Basic information gene, ability information gene, credit information gene and the posterior infromation gene answered.Later, above-mentioned various aspects are based on
Gene, constitute alternative interpreter's list of genes.For example, it is corresponding standby that basic information can be constructed for the basic information of interpreter
The person's of translating selectively list of genes is as shown in table 1, for according to a kind of alternative interpreter's list of genes of basic information of the embodiment of the present invention.
Table 1, a kind of alternative interpreter's list of genes of basic information according to an embodiment of the present invention
Then, when carrying out the selection of multiple interpreter's genomes according to table 1, multiple points in each data item can be randomly selected
Not corresponding interpreter's gene, such as selection to " institute learns profession " corresponding gene " oil exploitation " and " abroad works and learns to pass through
Go through " corresponding gene " having ", then interpreter's genome is constituted with the two.Using same treatment process, can also choose not
Multiple and different interpreter's genomes.
If can choose likewise, decimation rule has been previously set to choose gene relevant to interpreter's qualification information
The corresponding gene such as " IM ", " learning profession ", " date of birth " and " overseas work experience " in table 1 constitutes interpreter's genome.
The choosing method of interpreter's gene provided in an embodiment of the present invention, by from the basic information of interpreter, ability information, letter
With four aspects of information and posterior infromation, the gene of interpreter is extracted respectively, and constitute alternative interpreter's list of genes accordingly, to carry out
The selection and matching of more excellent interpreter's gene can more fully consider the specific information of interpreter's different aspect, for more reasonably into
The matching of row gene provides reliable basis.
Wherein, according to the above embodiments optionally, respectively from all basic informations of interpreter, ability information, credit letter
The step of extracting corresponding gene in breath and posterior infromation further comprises:
Obtain interpreter all basic informations, ability information, credit information and posterior infromation, and respectively from basic information,
Interpreter's feature is obtained in ability information, credit information and posterior infromation；
Based on interpreter's feature, interpreter's direct gene of interpreter is extracted.
It is to be understood that interpreter's gene is present in interpreter, different interpreters have different genes, there is general character but more important
Be otherness to be extracted gene, can just be treated in this way with differentiation, match best interpreter.
But gene is not feature, simply can not explicitly recognize, extract so needing step.Gene and spy
Sign is characterized in taking out a certain concept to characteristic common to object there are essential distinction.Include segment attribute in feature, and belongs to
Most basic information --- the gene of object included in property.
Therefore the present embodiment is when carrying out the extraction of interpreter's gene, first according to the four of the interpreter of above-described embodiment aspects
Information extracts corresponding characteristic information, as interpreter's feature.Later, according to different interpreter's features, interpreter is extracted most respectively
Basic information constitutes interpreter's direct gene.For example, as shown in Fig. 2, for according to the choosing of interpreter's gene provided in an embodiment of the present invention
Take the relation schematic diagram of interpreter's feature and interpreter's gene in method.
The choosing method of interpreter's gene provided in an embodiment of the present invention is further extracted by the extraction to interpreter's feature
Interpreter's gene, what can be got is present in specific interpreter, be different from other interpreters, unique key message.
Wherein, according to the above embodiments optionally, multiple matching result sampling is carried out, multiple successful match rate samples are obtained
This step of, further comprises:
Matching result sampling multiple for any wheel, executes following process flow:
The initial value of the successful match rate of all interpreter's genomes is initially set；
Interpreter's genome is randomly selected from all interpreter's genomes, and interpreter's genome of selection is matched
Test, and based on to interpreter's genome this match test successful match rate result and history match success rate as a result, more
The current successful match rate value of new interpreter's genome；
It repeats and randomly selects to the step of update, until the number to the match test of any interpreter's genome reaches
First given threshold stops the match test to interpreter's genome, and records the current successful match rate of interpreter's genome
Value；
To interpreter's genome other than the interpreter's genome for stopping match test, the step randomly selected to record is repeated
Suddenly, until reaching the second given threshold to the total degree of the match test of all interpreter's genomes, then each interpreter's gene is recorded
The current successful match rate value of group, and terminate the multiple matching result sampling of epicycle, into the multiple matching result sampling of next round, directly
Reach third given threshold to the total wheel number for executing multiple matching result sampling, the quantity for obtaining each interpreter's genome is third
The successful match rate sample of given threshold.
It is to be understood that according to the above embodiments, for interpreter's genome that each group selects, it is thus necessary to determine that
The matching effect of itself and contribution, so that selection is more suitable for the matched interpreter's gene of gene.Meanwhile in order to without loss of generality, for
Each group of interpreter's genome carries out multiple matching result sampling.And specifically carrying out each group of interpreter's genome selected
When matching result samples, carried out using above-mentioned Matching Model.
Specifically, can use given Matching Model, carry out the multiple matching results sampling of more wheels, obtain multiple matchings at
When power sample, it can be assumed that have chosen m group interpreter's genome according to the above embodiments, then it can be to each interpreter's genome
Successful match rate sampled, (no less than general 30 times) matching experiment is carried out repeatedly based on the above m genome, it is every to take turns
It is as follows with test process:
Step 1, initializing set, such as Initialize installation are carried out to the value of the successful match rate of each interpreter's genome
It is 0.
Step 2, interpreter's genome is randomly choosed, successful match rate result is carried out in given Matching Model and calculates,
Obtain the successful match rate result of this match test.Meanwhile in conjunction in epicycle multiple matching result sampling historical record it
The successful match rate of preceding match test for several times is as a result, i.e. history match success rate is as a result, calculate the interpreter's genome chosen
Current successful match rate value.
Step 3, repeatedly circulation executes above-mentioned steps 1 and 2, is from all interpreter's bases due to choosing interpreter's genome every time all
Because randomly selecting in group, therefore each genome may be different by the number of carry out match test, then when to some interpreter's gene
When the number of the match test of group reaches the first given threshold, that is, stop the epicycle match test to interpreter's genome,
And when recording stopping test, the current successful match rate value of interpreter's genome.
Step 4, remaining interpreter's genome except interpreter's genome of the first given threshold is reached for removing, continues to hold
The process flow of row above-mentioned steps 1-3 stops epicycle matching until the total degree of epicycle match test reaches the second given threshold
Test.At this point for each interpreter's genome, there is a successful match rate value to be corresponding to it, as the multiple matching result of epicycle
M successful match rate sample can be obtained then for m interpreter's genome by sampling obtained successful match rate sample.
So, for all interpreter's genomes, the above-mentioned multiple matching of more wheels (such as reaching third given threshold) is carried out
As a result sample, it can obtain multiple successful match rate samples of each interpreter's genome, such as wheel number is set as n, then matching at
Power sample number is n (n is typically no less than 50).
For example, it is assumed that having selected a1、a2And a3Totally three interpreter's genomes, and preset the first given threshold, second
Given threshold and third given threshold are respectively 3,8 and 5.Then, in the multiple matching result sampling of each round:
First time selection is carried out first, from a1、a2And a3In randomly select one, such as choose and arrive a1, then to a1Progress
With test, test result is successful match, then obtains a1Successful match rate value be 100%.
It is chosen followed by second, it is assumed that choose and arrive a2, match test is carried out to it, obtains test result as matching
It is unsuccessful, then obtain a2Successful match rate value be 0%.
Next third time selection is carried out again, it is assumed that and choose and arrive a1, and match test result is to match unsuccessful, then root
According to a1In total twice match test as a result, obtaining a1Current successful match rate value is 50%.
Next the 4th selection is carried out again, it is assumed that is chosen and is arrived a3, and match test result is successful match, then obtains a3
Successful match rate value be 100%.
Next the 5th selection is carried out again, it is assumed that and choose and arrive a1, and match test result is successful match, then basis
To a1In total three times match test as a result, obtaining a1Current successful match rate value is 66.6%.At this point, to a1Matching examination
It tests number and has had reached the first given threshold 3, then stop continuing to a1Carry out match test, and export its current matching at
Interpreter's genome a in the multiple matching result sampling of performance number 66.6%, as epicycle1Successful match rate sample.
Next the 6th selection is carried out again, due to a13 match tests are had reached, then only in a2And a3Middle progress
Match test is randomly selected and carries out, specific selection and match test process are similar with above-mentioned steps.In this way, until total matching
The number of test, i.e., to a1、a2And a3The total degree of match test when reaching the second given threshold 8 times, terminate epicycle multiple
It is sampled with result.At this point, having obtained a successful match rate sample all in accordance with above-mentioned match test for each interpreter's genome
This.
So, to three interpreter's genome a1、a2And a3, repeat more wheels and carry out above-mentioned multiple matching result sampling, then often
One wheel can obtain a1、a2And a3Corresponding one group of successful match rate sample.Until duplicate discussion reaches third given threshold
5, then available a1、a2And a3Respectively corresponding 5 successful match rate sample.
The choosing method of interpreter's gene provided in an embodiment of the present invention carries out each interpreter's genome using given Matching Model
Multiple successful match rate calculate, and accordingly choose the higher interpreter's genome of successful match rate, calculated result reliability can be made
It is higher.
Wherein, according to the above embodiments optionally, it is based on the corresponding mean value of interpreter's genome and standard deviation, and most
Big mean value and maximum standard deviation, the step of calculating interpreter's genome corresponding Z value, further comprise:
It using following calculation formula, calculates in all interpreter's genomes, each in addition to maximum interpreter's genome is translated
The corresponding Z value of member's genome:
In formula, ZiIndicate the corresponding Z value of i-th of interpreter's genome, n indicates the corresponding successful match of each interpreter's genome
The number of rate sample, EiIndicate the corresponding mean value of i-th of interpreter's genome, SiIndicate the corresponding standard of i-th of interpreter's genome
Difference, EmaxIndicate Largest Mean, SmaxIndicate maximum standard deviation.
It is to be understood that being translated in conjunction with each is calculated in addition to maximum interpreter's genome according to the above embodiments
The standard deviation and mean value of the corresponding successful match rate of member's genome, and the corresponding Largest Mean E of maximum interpreter's genomemaxWith
Maximum standard deviation Smax, using Z value calculation formula given herein above, can correspond to calculate select except maximum interpreter's gene
Each interpreter's genome Z value except group.
The choosing method of interpreter's gene provided in an embodiment of the present invention is distinguished using each the interpreter's genome selected
Corresponding mean value and standard deviation calculate the Z value of each interpreter's genome in combination with Largest Mean and maximum standard deviation, can
The successful match rate situation of each interpreter's genome is more accurately characterized, is come and original text so as to more accurately choose interpreter's gene
Part gene is matched, and matching effect is improved.
In addition, meeting translating for setting condition choosing from all interpreter's genomes on the basis of the various embodiments described above
After the step of member's genome, the method for the embodiment of the present invention can also include following processing step: if all except maximum interpreter
In interpreter's genome except genome, the Z value of none interpreter's genome can satisfy the setting condition of above-described embodiment,
Step S101 is then returned to, the different interpreter's genome of multiple groups is chosen again from alternative interpreter's list of genes, re-starts above-mentioned
The calculating and selection process of embodiment.
For example, the case where successful match rate sample for sampling meets normal distribution, to obtain 95% confidence level,
I.e. preset setting condition is that the confidence level of interpreter's genome meets 95%, then the Z value calculated for interpreter's genome is answered
No more than 1.96.And in practical application, it, may be due to being when choosing multiple groups interpreter genome from alternative interpreter's list of genes
It the reasons such as randomly selects, causes when calculating Z value to the interpreter's genome selected, Z value is not able to satisfy above-mentioned standard, then needs
Again other interpreter's genome is selected in alternative interpreter's list of genes, and is recalculated and chosen.
The choosing method of interpreter's gene provided in an embodiment of the present invention, by the judgement to calculated result and to selecting step
Circulating repetition execute, can guarantee that the high quality gene met the requirements can be selected, it is to be translated for more accurately matching
Contribution is of great significance.
As the other side of the embodiment of the present invention, the embodiment of the present invention provides a kind of interpreter according to the above embodiments
The selecting device of gene, the device are used to realize the selection to final interpreter's gene in the above embodiments.Therefore, above-mentioned
Description and definition in the choosing method of interpreter's gene of each embodiment can be used for each execution module in the embodiment of the present invention
Understanding, specifically refer to above-described embodiment, do not repeating herein.
One embodiment of present aspect embodiment according to the present invention, the structure of the selecting device of interpreter's gene as shown in figure 3,
For the structural schematic diagram of the selecting device of interpreter's gene provided in an embodiment of the present invention, which can be used for above-mentioned each method
The selection of interpreter's gene in embodiment, the device include: that initial gene chooses module 301, the first computing module 302, maximum base
Because a group selection module 303, the second computing module 304 and final gene choose module 305.
Wherein, initial gene is chosen module 301 and is used for from alternative interpreter's list of genes, chooses the different base of multiple groups respectively
Cause constitutes multiple interpreter's genomes；First computing module 302 is used to carry out repeatedly matching knot for each interpreter's genome
Fruit sampling obtains multiple successful match rate samples, and is based on multiple successful match rate samples, and it is corresponding to calculate interpreter's genome
The mean value and standard deviation of successful match rate；It is corresponding that maximum genome chooses the maximum that module 303 is used to choose in all mean values
Interpreter's genome, be defined as maximum interpreter's genome, and the mean value of maximum interpreter's genome is defined as Largest Mean, will most
The standard deviation of big interpreter's genome is defined as maximum standard deviation；Second computing module 304 is used for in all interpreter's genomes
Each interpreter's genome in addition to maximum interpreter's genome is based on the corresponding mean value of interpreter's genome and standard deviation, with
And Largest Mean and maximum standard deviation, calculate the corresponding Z value of interpreter's genome；Final gene chooses module 305 for being based on
The corresponding Z value of each interpreter's genome in all interpreter's genomes in addition to maximum interpreter's genome, from all interpreter's bases
Because choosing the interpreter's genome for meeting and imposing a condition in group, and gene in interpreter's genome that the satisfaction is imposed a condition and most
Gene in big interpreter's genome merges, and obtains the interpreter's gene finally chosen；Wherein, the Z value indicates large sample otherness
Z value in verifying.
Specifically, initial gene chooses module 301 can select respectively according to the alternative interpreter's list of genes pre-established
Multiple groups interpreter's gene is taken, and a genome is constituted with each group of interpreter's gene respectively, as interpreter's genome, interpreter's gene
Group is the interpreter's genome selected.For example, initial gene chooses module 301 can when carrying out the selection of each group interpreter gene
With from multiple interpreter's genes in alternative interpreter's list of genes in random selection table, and the interpreter's gene randomly selected using these
Constitute a genome, as interpreter's genome.
Later, for the interpreter's genome selected for each group, it is thus necessary to determine that the matching effect of itself and contribution, thus
Selection is more suitable for the matched interpreter's gene of gene.Meanwhile in order to without loss of generality, be counted for each group of interpreter's genome, first
Calculating module 302 can repeatedly be matched by the way that interpreter's genome is inputted given Matching Model using given Matching Model
As a result it samples, sampling can obtain a successful match rate sample every time.It is understood that each successful match rate sample
This, the successful match rate score that an actually matching result samples.
In addition, the first computing module 302 is according to above-mentioned multiple matching for each interpreter selected genome
As a result the multiple successful match rate samples for sampling acquisition, calculate the comprehensive matching success rate of interpreter's genome, that is, calculate separately
The mean value and standard deviation of the corresponding successful match rate of interpreter's genome.
Later, maximum genome chooses module 303 and chooses value the maximum first from all mean values of above-mentioned calculating, and
Interpreter's genome corresponding to the maximum is defined as maximum interpreter's genome.Then correspondingly, maximum genome chooses module
The mean value of maximum interpreter's genome is also defined as Largest Mean by 303, with variable EmaxIt indicates, by maximum interpreter's genome
Standard deviation be defined as maximum standard deviation, with variable SmaxIt indicates.
Later, the second computing module 304 is according to the above-mentioned all mean values and standard deviation being calculated, maximum gene at calculating
The Z value of each genome except group.Specifically, for each of above-mentioned interpreter's genome interpreter's genome, the second meter
The standard deviation and mean value for calculating successful match rate of the module 304 according to corresponding to it, in conjunction with the corresponding maximum of maximum interpreter's genome
Mean value and maximum standard deviation calculate separately the corresponding Z value of interpreter's genome.
Finally, may determine that each corresponding interpreter's genome according to the Z value of each interpreter's genome on the basis of above-mentioned calculating
Otherness performance when carrying out gene matching.Therefore, according to the corresponding Z value of each interpreter's genome, final gene chooses mould
Block 305 can use preset setting condition, judge whether the corresponding interpreter's genome of the Z value meets the otherness of setting
It is required that.If conditions are not met, then reject it from each interpreter's genome selected, what final residue was not removed all is translated
Member's genome is satisfactory interpreter's genome, includes that Z value meets setting otherness requirement in these interpreter's genomes
Interpreter's genome and maximum interpreter's genome.Finally, final gene selection module 305 will be in remaining all interpreter's genomes
Gene takes out, and after removing the duplicate factor in these genes, forms one group of new gene, i.e., as the interpreter finally chosen
Gene.
Further, on the basis of the above embodiments, the device of the embodiment of the present invention further includes alternative interpreter's gene column
Table constructs module, is used for: extracting phase from all basic informations of interpreter, ability information, credit information and posterior infromation respectively
The gene answered, and it is correspondingly formed basic information gene, ability information gene, credit information gene and the posterior infromation base of interpreter
Cause；Based on the basic information gene, ability information gene, credit information gene and posterior infromation gene, constitute described alternative
Interpreter's list of genes.
Wherein optional, alternative interpreter's list of genes building module is specifically used for: obtaining all basic informations, the energy of interpreter
Force information, credit information and posterior infromation, and obtained from basic information, ability information, credit information and posterior infromation respectively
Interpreter's feature；Based on interpreter's feature, interpreter's direct gene of interpreter is extracted.
Wherein optional, the second computing module is specifically used for: utilizing following formula, calculates in all interpreter's genomes, remove
The corresponding Z value of each interpreter's genome except maximum interpreter's genome:
In formula, ZiIndicate the corresponding Z value of i-th of interpreter's genome, n indicates the corresponding successful match of each interpreter's genome
The number of rate sample, EiIndicate the corresponding mean value of i-th of interpreter's genome, SiIndicate the corresponding standard of i-th of interpreter's genome
Difference, EmaxIndicate Largest Mean, SmaxIndicate maximum standard deviation.
Wherein optional, the first computing module is specifically used for: any multiple matching result of wheel being sampled, following place is executed
Reason process: the initial value of the successful match rate of all interpreter's genomes is initially set；From all interpreter's genomes with
Machine chooses interpreter's genome, carries out match test to interpreter's genome of selection, and based on to interpreter's genome sheet
The successful match rate result and history match success rate of secondary match test are as a result, update the current successful match of interpreter's genome
Rate value；It repeats and randomly selects to the step of update, until reaching the to the number of the match test of any interpreter's genome
One given threshold stops the match test to interpreter's genome, and records the current successful match rate value of interpreter's genome；
To stop match test interpreter's genome other than interpreter's genome, repeat randomly select to record the step of, until
Second given threshold is reached to the total degree of the match test of all interpreter's genomes, then it is current to record each interpreter's genome
Successful match rate value, and terminate the multiple matching result sampling of epicycle, into the multiple matching result sampling of next round, until executing more
Total wheel number of secondary matching result sampling reaches third given threshold, and the quantity for obtaining each interpreter's genome is third given threshold
Successful match rate sample.
Wherein optional, the first computing module is specifically used for, for each interpreter's genome, the successful match rate of extraction
The number of sample is no less than given threshold.
It is understood that can be by hardware processor (hardware processor) come real in the embodiment of the present invention
Each relative program module in the device of existing the various embodiments described above.Also, the selecting device of each interpreter's gene of the embodiment of the present invention
When for the selection of interpreter's gene in above-mentioned each method embodiment, the beneficial effect of generation and corresponding above-mentioned each method are real
It is identical to apply example, above-mentioned each method embodiment can be referred to, details are not described herein again.
As the another aspect of the embodiment of the present invention, the present embodiment provides a kind of electronics according to the above embodiments and sets
It is standby, it is the entity structure schematic diagram of electronic equipment provided in an embodiment of the present invention, comprising: at least one processor with reference to Fig. 4
401, at least one processor 402, communication interface 403 and bus 404.
Wherein, memory 401, processor 402 and communication interface 403 complete mutual communication by bus 404, communicate
Interface 403 is for the information transmission between the electronic equipment and interpreter's information equipment；Being stored in memory 401 can be in processor
The computer program run on 402 when processor 402 executes the computer program, realizes translating as described in the various embodiments described above
The choosing method of member's gene.
It is to be understood that including at least memory 401, processor 402, communication interface 403 and bus in the electronic equipment
404, and memory 401, processor 402 and communication interface 403 form mutual communication connection by bus 404, and can be complete
The program instruction of the choosing method of interpreter's gene is read from memory 401 at mutual communication, such as processor 402.Separately
Outside, communication interface 403 can also realize the communication connection between the electronic equipment and interpreter's information equipment, and achievable mutual
Information transmission, such as the selection to interpreter's gene is realized by communication interface 403.
When electronic equipment is run, processor 402 calls the program instruction in memory 401, real to execute above-mentioned each method
Apply method provided by example, for example, from alternative interpreter's list of genes, choose the different gene of multiple groups respectively, constitute more
A interpreter's genome；For each interpreter's genome, multiple matching result sampling is carried out, obtains multiple successful match rate samples
This, and multiple successful match rate samples are based on, calculate the mean value and standard deviation of the corresponding successful match rate of interpreter's genome；Choosing
The corresponding interpreter's genome of the maximum in all mean values is taken, is defined as maximum interpreter's genome, and by maximum interpreter's genome
Mean value be defined as Largest Mean, the standard deviation of maximum interpreter's genome is defined as maximum standard deviation；For all interpreter's bases
Because of each interpreter's genome in group in addition to maximum interpreter's genome, it is based on the corresponding mean value of interpreter's genome and standard
Difference and Largest Mean and maximum standard deviation calculate the corresponding Z value of interpreter's genome；Based on being removed in all interpreter's genomes
The corresponding Z value of each interpreter's genome except maximum interpreter's genome is chosen from all interpreter's genomes and meets setting
Interpreter's genome of condition, and the base in the gene in interpreter's genome that the satisfaction is imposed a condition and maximum interpreter's genome
Because merging, the interpreter's gene finally chosen is obtained；Wherein, the Z value indicates Z value etc. in the verifying of large sample otherness.
Program instruction in above-mentioned memory 401 can be realized and as independent by way of SFU software functional unit
Product when selling or using, can store in a computer readable storage medium.Alternatively, realizing that above-mentioned each method is implemented
This can be accomplished by hardware associated with program instructions for all or part of the steps of example, and program above-mentioned can store to be calculated in one
In machine read/write memory medium, when being executed, execution includes the steps that above-mentioned each method embodiment to the program；And storage above-mentioned
Medium includes: USB flash disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random
Access Memory, RAM), the various media that can store program code such as magnetic or disk.
The embodiment of the present invention provides a kind of non-transient computer readable storage medium also according to the various embodiments described above, this is non-temporarily
State computer-readable recording medium storage computer instruction, the computer instruction execute computer as described in the various embodiments described above
Interpreter's gene choosing method, for example, from alternative interpreter's list of genes, choose the different gene of multiple groups, structure respectively
At multiple interpreter's genomes；For each interpreter's genome, multiple matching result sampling is carried out, obtains multiple successful match rates
Sample, and multiple successful match rate samples are based on, calculate the mean value and standard deviation of the corresponding successful match rate of interpreter's genome；
The corresponding interpreter's genome of the maximum in all mean values is chosen, is defined as maximum interpreter's genome, and by maximum interpreter's gene
The mean value of group is defined as Largest Mean, and the standard deviation of maximum interpreter's genome is defined as maximum standard deviation；For all interpreters
Each interpreter's genome in genome in addition to maximum interpreter's genome is based on the corresponding mean value of interpreter's genome and mark
Quasi- poor and Largest Mean and maximum standard deviation, calculate the corresponding Z value of interpreter's genome；Based in all interpreter's genomes
The corresponding Z value of each interpreter's genome in addition to maximum interpreter's genome is chosen satisfaction from all interpreter's genomes and is set
Interpreter's genome of fixed condition, and in the gene in interpreter's genome that the satisfaction is imposed a condition and maximum interpreter's genome
Gene merges, and obtains the interpreter's gene finally chosen；Wherein, the Z value indicates Z value etc. in the verifying of large sample otherness.
Electronic equipment provided in an embodiment of the present invention and non-transient computer readable storage medium, by executing above-mentioned each reality
The choosing method of interpreter's gene described in example is applied, chooses multiple groups interpreter genome from interpreter's gene pool of all interpreters in advance,
And by calculate these interpreter's genomes corresponding to Z value, come choose Z value meet impose a condition interpreter's genome, using as
Final chooses as a result, the interpreter's gene selected is enabled preferably to embody the otherness between interpreter.In addition, in gene
With in application, the interpreter chosen accordingly can be made more reasonably to be matched with contribution to be translated, to effectively improve translation effect
Rate and translation accuracy rate.
It is understood that the embodiment of device described above, electronic equipment and storage medium is only schematic
, wherein unit may or may not be physically separated as illustrated by the separation member, it can both be located at one
Place, or may be distributed on heterogeneous networks unit.Some or all of modules can be selected according to actual needs
To achieve the purpose of the solution of this embodiment.Those of ordinary skill in the art are without paying creative labor
To understand and implement.
By the description of embodiment of above, those skilled in the art is it will be clearly understood that each embodiment can borrow
Help software that the mode of required general hardware platform is added to realize, naturally it is also possible to pass through hardware.Based on this understanding, above-mentioned
Substantially the part that contributes to existing technology can be embodied in the form of software products technical solution in other words, the meter
Calculation machine software product may be stored in a computer readable storage medium, such as USB flash disk, mobile hard disk, ROM, RAM, magnetic disk or light
Disk etc., including some instructions, with so that a computer equipment (such as personal computer, server or network equipment etc.)
Execute method described in certain parts of above-mentioned each method embodiment or embodiment of the method.
In addition, those skilled in the art are it should be understood that in the application documents of the embodiment of the present invention, term
"include", "comprise" or any other variant thereof is intended to cover non-exclusive inclusion, so that including a series of elements
Process, method, article or equipment not only include those elements, but also including other elements that are not explicitly listed, or
Person is to further include for elements inherent to such a process, method, article, or device.In the absence of more restrictions, by
The element that sentence "including a ..." limits, it is not excluded that in the process, method, article or apparatus that includes the element
There is also other identical elements.
In the specification of the embodiment of the present invention, numerous specific details are set forth.It should be understood, however, that the present invention is implemented
The embodiment of example can be practiced without these specific details.In some instances, it is not been shown in detail well known
Methods, structures and technologies, so as not to obscure the understanding of this specification.Similarly, it should be understood that in order to simplify implementation of the present invention
Example is open and helps to understand one or more of the various inventive aspects, above to the exemplary embodiment of the embodiment of the present invention
Description in, each feature of the embodiment of the present invention is grouped together into single embodiment, figure or descriptions thereof sometimes
In.
However, the disclosed method should not be interpreted as reflecting the following intention: i.e. the claimed invention is implemented
Example requires features more more than feature expressly recited in each claim.More precisely, such as claims institute
As reflection, inventive aspect is all features less than single embodiment disclosed above.Therefore, it then follows specific embodiment party
Thus claims of formula are expressly incorporated in the specific embodiment, wherein each claim itself is real as the present invention
Apply the separate embodiments of example.
Finally, it should be noted that above embodiments are only to illustrate the technical solution of the embodiment of the present invention, rather than it is limited
System；Although the embodiment of the present invention is described in detail with reference to the foregoing embodiments, those skilled in the art it is understood that
It is still possible to modify the technical solutions described in the foregoing embodiments, or part of technical characteristic is carried out etc.
With replacement；And these are modified or replaceed, each embodiment skill of the embodiment of the present invention that it does not separate the essence of the corresponding technical solution
The spirit and scope of art scheme.
Claims (8)
1. a kind of choosing method of interpreter's gene characterized by comprising
From alternative interpreter's list of genes, the different gene of multiple groups is chosen respectively, constitutes multiple interpreter's genomes；
For interpreter's genome described in each, multiple matching result sampling is carried out, obtains multiple successful match rate samples, and base
In the multiple successful match rate sample, the mean value and standard deviation of the corresponding successful match rate of interpreter's genome are calculated；
The corresponding interpreter's genome of the maximum in all mean values is chosen, is defined as maximum interpreter's genome, and will be described
The mean value of maximum interpreter's genome is defined as Largest Mean, and the standard deviation of maximum interpreter's genome is defined as
Maximum standard deviation；
For each described interpreter's genome in all interpreter's genomes in addition to maximum interpreter's genome, base
In the corresponding mean value of interpreter's genome and the standard deviation and the Largest Mean and the maximum standard deviation, meter
Calculate the corresponding Z value of interpreter's genome；
Based on each described interpreter's genome pair in all interpreter's genomes in addition to maximum interpreter's genome
The Z value answered chooses the interpreter's genome for meeting and imposing a condition from all interpreter's genomes, and the satisfaction is set
The gene in gene and maximum interpreter's genome in interpreter's genome of fixed condition merges, and obtains the interpreter finally chosen
Gene；
Wherein, the Z value indicates Z value in the verifying of large sample otherness.
2. the method according to claim 1, wherein being chosen respectively from alternative interpreter's list of genes described
Before the step of multiple groups different gene, further includes:
Corresponding gene is extracted from all basic informations of interpreter, ability information, credit information and posterior infromation respectively, and right
Basic information gene, ability information gene, credit information gene and the posterior infromation gene of interpreter should be formed；
Based on the basic information gene, ability information gene, credit information gene and posterior infromation gene, constitute described alternative
Interpreter's list of genes.
3. the method according to claim 1, wherein it is described based on the corresponding mean value of interpreter's genome and
The standard deviation and the Largest Mean and the maximum standard deviation, the step of calculating interpreter's genome corresponding Z value into
One step includes:
Using following calculation formula, calculate in all interpreter's genomes, it is each in addition to maximum interpreter's genome
The corresponding Z value of a interpreter's genome:
In formula, ZiIndicate the corresponding Z value of i-th of interpreter's genome, n indicates the corresponding matching of each interpreter's genome
The number of success rate sample, EiIndicate the corresponding mean value of i-th of interpreter's genome, SiIndicate that i-th of interpreter's genome is corresponding
The standard deviation, EmaxIndicate the Largest Mean, SmaxIndicate the maximum standard deviation.
4. obtaining multiple the method according to claim 1, wherein described carry out multiple matching result sampling
The step of being made into power sample further comprises:
The multiple matching result sampling described for any wheel, executes following process flow:
The initial value of the successful match rate of all interpreter's genomes is initially set；
Interpreter's genome is randomly selected from all interpreter's genomes, and interpreter's genome of selection is carried out
Match test, and based on the successful match rate result and history match success rate knot to this match test of interpreter's genome
Fruit updates the current successful match rate value of interpreter's genome；
Repeat it is described randomly select to the step of the update, until to the match test of any interpreter's genome
Number reaches the first given threshold, stops the match test to interpreter's genome, and records current of interpreter's genome
With success ratio values；
To interpreter's genome other than the interpreter's genome for stopping match test, described randomly select to the record is repeated
The step of, until reaching the second given threshold to the total degree of the match test of all interpreter's genomes, then record is each
The current successful match rate value of interpreter's genome, and terminate multiple matching result sampling described in epicycle, into next round institute
Multiple matching result sampling is stated, until the total wheel number for executing the multiple matching result sampling reaches third given threshold, is obtained
The quantity of each interpreter's genome is the successful match rate sample of third given threshold.
5. the method according to claim 1, wherein for interpreter's genome described in each, extraction it is described
The number of successful match rate sample is no less than given threshold.
6. a kind of selecting device of interpreter's gene characterized by comprising
Initial gene chooses module, for choosing the different gene of multiple groups respectively, constituting multiple from alternative interpreter's list of genes
Interpreter's genome；
First computing module obtains multiple for carrying out multiple matching result sampling for interpreter's genome described in each
It is made into power sample, and is based on the multiple successful match rate sample, calculates the corresponding successful match rate of interpreter's genome
Mean value and standard deviation；
Maximum genome is chosen module and is defined as choosing the corresponding interpreter's genome of the maximum in all mean values
Maximum interpreter's genome, and the mean value of maximum interpreter's genome is defined as Largest Mean, by the maximum interpreter
The standard deviation of genome is defined as maximum standard deviation；
Second computing module, for for each in all interpreter's genomes in addition to maximum interpreter's genome
Interpreter's genome, based on the corresponding mean value of interpreter's genome and the standard deviation and the Largest Mean and
The maximum standard deviation calculates the corresponding Z value of interpreter's genome；
Final gene chooses module, for based on every in addition to maximum interpreter's genome in all interpreter's genomes
The corresponding Z value of one interpreter's genome, chooses the interpreter for meeting and imposing a condition from all interpreter's genomes
Genome, and the gene in the gene and maximum interpreter's genome in the interpreter's genome for meeting and imposing a condition is closed
And obtain the interpreter's gene finally chosen；
Wherein, the Z value indicates Z value in the verifying of large sample otherness.
7. a kind of electronic equipment characterized by comprising at least one processor, at least one processor, communication interface and total
Line；
The memory, the processor and the communication interface complete mutual communication, the communication by the bus
Interface is for the information transmission between the electronic equipment and interpreter's information equipment；
The computer program that can be run on the processor is stored in the memory, the processor executes the calculating
When machine program, the method as described in any in claim 1 to 5 is realized.
8. a kind of non-transient computer readable storage medium, which is characterized in that the non-transient computer readable storage medium is deposited
Computer instruction is stored up, the computer instruction makes the computer execute the method as described in any in claim 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811096578.6A CN109448792B (en) | 2018-09-19 | 2018-09-19 | Translator gene selection method and device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811096578.6A CN109448792B (en) | 2018-09-19 | 2018-09-19 | Translator gene selection method and device and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109448792A true CN109448792A (en) | 2019-03-08 |
| CN109448792B CN109448792B (en) | 2021-11-05 |
Family
ID=65532883
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811096578.6A Active CN109448792B (en) | 2018-09-19 | 2018-09-19 | Translator gene selection method and device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109448792B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101853398A (en) * | 2010-05-11 | 2010-10-06 | 浙江大学 | Chinese paper cutting identification method based on space constraint characteristic selection and combination thereof |
| CN101937440A (en) * | 2009-06-30 | 2011-01-05 | 华为技术有限公司 | Feature selection method and device |
| CN102080129A (en) * | 2010-12-01 | 2011-06-01 | 杭州师范大学 | Multiple method-based differential gene screening control method for gene chip |
| CN103064970A (en) * | 2012-12-31 | 2013-04-24 | 武汉传神信息技术有限公司 | Search method for optimizing translators |
| CN106326458A (en) * | 2016-06-02 | 2017-01-11 | 广西智度信息科技有限公司 | Method for classifying city management cases based on text classification |
| CN107463799A (en) * | 2017-08-23 | 2017-12-12 | 福建师范大学福清分校 | Interaction fusion feature represents the DBP recognition methods with selective ensemble |
| CN107808100A (en) * | 2017-10-25 | 2018-03-16 | 中国科学技术大学 | For the steganalysis method of fc-specific test FC sample |
| CN107885730A (en) * | 2017-09-25 | 2018-04-06 | 沈阳航空航天大学 | Translation knowledge method for distinguishing validity under more interpreter's patterns |
-
2018
- 2018-09-19 CN CN201811096578.6A patent/CN109448792B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101937440A (en) * | 2009-06-30 | 2011-01-05 | 华为技术有限公司 | Feature selection method and device |
| CN101853398A (en) * | 2010-05-11 | 2010-10-06 | 浙江大学 | Chinese paper cutting identification method based on space constraint characteristic selection and combination thereof |
| CN102080129A (en) * | 2010-12-01 | 2011-06-01 | 杭州师范大学 | Multiple method-based differential gene screening control method for gene chip |
| CN103064970A (en) * | 2012-12-31 | 2013-04-24 | 武汉传神信息技术有限公司 | Search method for optimizing translators |
| CN106326458A (en) * | 2016-06-02 | 2017-01-11 | 广西智度信息科技有限公司 | Method for classifying city management cases based on text classification |
| CN107463799A (en) * | 2017-08-23 | 2017-12-12 | 福建师范大学福清分校 | Interaction fusion feature represents the DBP recognition methods with selective ensemble |
| CN107885730A (en) * | 2017-09-25 | 2018-04-06 | 沈阳航空航天大学 | Translation knowledge method for distinguishing validity under more interpreter's patterns |
| CN107808100A (en) * | 2017-10-25 | 2018-03-16 | 中国科学技术大学 | For the steganalysis method of fc-specific test FC sample |
Non-Patent Citations (4)
| Title |
|---|
| K.Z.MAO: ""Identifying Critical Variables of Principal Components for Unsupervised Feature Selection"", 《IEEE TRANSACTIONS ON SYSTEMS,MAN,AND CYBERNETICS-PART B:CYBERNETICS》 * |
| 亓呈明等: "《机器学习智能计算与高光谱遥感影像分类应用研究》", 31 May 2018 * |
| 程涛: ""基于机器学习的蛋白质相互作用预测结果与样本重复性关系的研究"", 《中国优秀硕士学位论文全文数据库 基础科学及》 * |
| 陈飞飞: ""基于特征表示的行为识别方法研究"", 《中国博士学位论文全文数据库 信息科技辑》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109448792B (en) | 2021-11-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103577989B (en) | A kind of information classification approach and information classifying system based on product identification | |
| CN108509408A (en) | A kind of sentence similarity judgment method | |
| CN112667794A (en) | Intelligent question-answer matching method and system based on twin network BERT model | |
| CN108509411A (en) | Semantic analysis and device | |
| CN110427461A (en) | Intelligent answer information processing method, electronic equipment and computer readable storage medium | |
| CN111309887B (en) | Method and system for training text key content extraction model | |
| CN105868179B (en) | A kind of intelligent answer method and device | |
| CN108804677A (en) | In conjunction with the deep learning question classification method and system of multi-layer attention mechanism | |
| CN106296195A (en) | A kind of Risk Identification Method and device | |
| CN110008327A (en) | Law answers generation method and device | |
| CN107273406A (en) | Dialog process method and device in task dialogue system | |
| CN108304364A (en) | keyword extracting method and device | |
| CN108804526A (en) | Interest determines that system, interest determine method and storage medium | |
| CN109857846A (en) | The matching process and device of user's question sentence and knowledge point | |
| CN111694937A (en) | Interviewing method and device based on artificial intelligence, computer equipment and storage medium | |
| CN113761218A (en) | Entity linking method, device, equipment and storage medium | |
| CN110390107A (en) | Hereafter relationship detection method, device and computer equipment based on artificial intelligence | |
| CN115687925A (en) | Fault type identification method and device for unbalanced sample | |
| CN114490998B (en) | Text information extraction method and device, electronic equipment and storage medium | |
| CN109918681A (en) | It is a kind of based on Chinese character-phonetic fusion problem semantic matching method | |
| CN109189892A (en) | A kind of recommended method and device based on article review | |
| CN110688478A (en) | Answer sorting method, device and storage medium | |
| TW201820172A (en) | System, method and non-transitory computer readable storage medium for conversation analysis | |
| CN115130538A (en) | Training method of text classification model, text processing method, equipment and medium | |
| CN115048505A (en) | Corpus screening method and device, electronic equipment and computer readable medium |
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 |