CN109785112A - Repayment method, computer readable storage medium and server neural network based - Google Patents
Repayment method, computer readable storage medium and server neural network based Download PDFInfo
- Publication number
- CN109785112A CN109785112A CN201811533834.3A CN201811533834A CN109785112A CN 109785112 A CN109785112 A CN 109785112A CN 201811533834 A CN201811533834 A CN 201811533834A CN 109785112 A CN109785112 A CN 109785112A
- Authority
- CN
- China
- Prior art keywords
- refunded
- task
- account
- savings account
- refund
- 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
- 238000000034 method Methods 0.000 title claims abstract description 32
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 7
- 238000003062 neural network model Methods 0.000 claims abstract description 29
- 230000004044 response Effects 0.000 claims description 13
- 238000005516 engineering process Methods 0.000 abstract description 3
- 230000006870 function Effects 0.000 description 9
- 238000012545 processing Methods 0.000 description 8
- 230000008569 process Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 4
- 230000001537 neural effect Effects 0.000 description 4
- 238000013475 authorization Methods 0.000 description 3
- 238000012790 confirmation Methods 0.000 description 3
- 102100029469 WD repeat and HMG-box DNA-binding protein 1 Human genes 0.000 description 2
- 101710097421 WD repeat and HMG-box DNA-binding protein 1 Proteins 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- PEDCQBHIVMGVHV-UHFFFAOYSA-N Glycerine Chemical compound OCC(O)CO PEDCQBHIVMGVHV-UHFFFAOYSA-N 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000010485 coping Effects 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000004218 nerve net Anatomy 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
Landscapes
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The invention belongs to field of computer technology more particularly to a kind of repayment methods neural network based, computer readable storage medium and server.The method obtains the information aggregate sent by specified data source first, it include the information of one or more in the information aggregate, then, task to be refunded is filtered out from the information aggregate according to preset keyword set, it include more than one keyword in the keyword set, finally, handling using preset neural network model the task to be refunded, the refund scheme to the task to be refunded is obtained.Through the embodiment of the present invention, it may be implemented to filter out task to be refunded from massive information automatically, and refund scheme be generated for user automatically, can avoid greatly improving user experience because carelessness causes to forget the case where refunding according to the refund scheme.
Description
Technical field
The invention belongs to field of computer technology more particularly to a kind of repayment method neural network based, computer can
Read storage medium and server.
Background technique
In actual life, many users possess a large amount of credit card and other bills for needing to refund on time, if not
It can refund in time, stain can be left to the credit record of user.In existing situation, generally requires user and voluntarily refund, but
It is that, in the case where the bill for needing to refund is more, user is likely to because carelessness causes to forget the case where refunding.
Summary of the invention
In view of this, the embodiment of the invention provides a kind of repayment methods neural network based, computer-readable storage
Medium and server, to solve in the case where the bill for needing to refund is more, user is likely to because carelessness causes to forget
The problem of the case where refund, occurs.
The first aspect of the embodiment of the present invention provides a kind of repayment method, may include:
The information aggregate sent by specified data source is obtained, includes the information of one or more in the information aggregate;
Task to be refunded is filtered out from the information aggregate according to preset keyword set, in the keyword set
Including more than one keyword;
The task to be refunded is handled using preset neural network model, is obtained to the task to be refunded
Refund scheme.
The second aspect of the embodiment of the present invention provides a kind of computer readable storage medium, the computer-readable storage
Media storage has computer-readable instruction, and the computer-readable instruction realizes following steps when being executed by processor:
The information aggregate sent by specified data source is obtained, includes the information of one or more in the information aggregate;
Task to be refunded is filtered out from the information aggregate according to preset keyword set, in the keyword set
Including more than one keyword;
The task to be refunded is handled using preset neural network model, is obtained to the task to be refunded
Refund scheme.
The third aspect of the embodiment of the present invention provides a kind of server, including memory, processor and is stored in institute
The computer-readable instruction that can be run in memory and on the processor is stated, the processor executes described computer-readable
Following steps are realized when instruction:
The information aggregate sent by specified data source is obtained, includes the information of one or more in the information aggregate;
Task to be refunded is filtered out from the information aggregate according to preset keyword set, in the keyword set
Including more than one keyword;
The task to be refunded is handled using preset neural network model, is obtained to the task to be refunded
Refund scheme.
Existing beneficial effect is the embodiment of the present invention compared with prior art: the embodiment of the present invention is obtained by specifying first
The information aggregate that sends of data source, include the information of one or more in the information aggregate, then, according to preset keyword
Set filters out task to be refunded from the information aggregate, includes more than one keyword in the keyword set, most
Afterwards, the task to be refunded is handled using preset neural network model, obtains the refund to the task to be refunded
Scheme.Through the embodiment of the present invention, it may be implemented to filter out task to be refunded from massive information automatically, and automatically raw for user
At refund scheme, can avoid greatly improving use because carelessness causes to forget the case where refunding according to the refund scheme
Family experience.
Detailed description of the invention
It to describe the technical solutions in the embodiments of the present invention more clearly, below will be to embodiment or description of the prior art
Needed in attached drawing be briefly described, it should be apparent that, the accompanying drawings in the following description is only of the invention some
Embodiment for those of ordinary skill in the art without any creative labor, can also be according to these
Attached drawing obtains other attached drawings.
Fig. 1 is a kind of one embodiment flow chart of repayment method in the embodiment of the present invention;
Fig. 2 is to treat refund task using preset neural network model to be handled, and obtains treating going back for refund task
The schematic flow diagram of money scheme;
Fig. 3 is the schematic diagram of neural network model used in the embodiment of the present invention;
Fig. 4 is a kind of one embodiment structure chart of refund device in the embodiment of the present invention;
Fig. 5 is a kind of schematic block diagram of server in the embodiment of the present invention.
Specific embodiment
In order to make the invention's purpose, features and advantages of the invention more obvious and easy to understand, below in conjunction with the present invention
Attached drawing in embodiment, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that disclosed below
Embodiment be only a part of the embodiment of the present invention, and not all embodiment.Based on the embodiments of the present invention, this field
Those of ordinary skill's all other embodiment obtained without making creative work, belongs to protection of the present invention
Range.
Referring to Fig. 1, a kind of one embodiment of repayment method may include: in the embodiment of the present invention
Step S101, the information aggregate sent by specified data source is obtained.
The data source includes the server of each bank or financial institution, includes one or more in the information aggregate
Information, these information are either short message, the PUSH message being also possible in application program (APP).
Step S102, task to be refunded is filtered out from the information aggregate according to preset keyword set.
Include more than one keyword in the keyword set, setting up procedure may include steps of:
Firstly, constructing the first corpus and the second corpus according to the historical information in preset database.
Include each history refund information in first corpus, includes letter of refunding except history in second corpus
Each other historical informations except breath.
Then, the frequency and each candidate word that each candidate word occurs in first corpus are counted respectively
The frequency that language occurs in second corpus.
Candidate's word is the word occurred in first corpus.
Then, the discrimination of each candidate word is calculated separately according to the following formula:
Wherein, w is the serial number of each candidate word, and 1≤w≤WordNum, WordNum are the sum of candidate word,
FstNumwFor the frequency that w-th of candidate word occurs in first corpus, SndNumwIt is w-th of candidate word in institute
State the frequency occurred in the second corpus, DivDegwFor the discrimination of w-th of candidate word.
Finally, selecting keyword, and the keyword structure that will be selected from each candidate word according to the discrimination
It makes as the keyword set.
For example, each candidate word can be arranged according to the sequence of discrimination from big to small, therefrom choose sequence near
Several preceding candidate words are the keyword of refund information, and the number of keyword can be configured according to the actual situation,
For example, 5,10,20 or other values can be set to.
It, can be according to the keyword set from the information aggregate under the premise of constructing the keyword set
Filter out task to be refunded.
If desired judge whether a certain information is refund information, then counts and contain how many a keywords in the information, if
Number comprising keyword is greater than threshold value, then determines that it is refund information, if the number comprising keyword is less than or equal to threshold value,
Then determining it not is refund information.
The threshold value can be specifically arranged according to the following method:
Firstly, calculating the examination accuracy under the various values of the threshold value according to the following formula:
Wherein, the value of the kn threshold value, 1≤kn≤KwNum, KwNum are the sum of keyword, FstPosNumknIt will to work as
The number of the refund information screened out from the first corpus when threshold value is set as kn, FstNegNumknIt is set a threshold to work as
The number of the non-refund information screened out from the first corpus when kn, SndPosNumknKn Shi Cong is set a threshold to work as
The number of the refund information screened out in two corpus, SndNegNumknFor when setting a threshold to kn from the second corpus
The number of the non-refund information screened out, DivAccknFor the examination accuracy when setting a threshold to kn.
Then, the value of the threshold value is determined according to the following formula:
BestNum=Argmax (DivAccSet)
=Argmax ([DivAcc1,DivAcc2,......,DivAcckn,......,DivAccKwNum])
Wherein, BestNum is the value of the threshold value, and Argmax is maximum independent variable function, and DivAccSet is the threshold value
Examination accuracy set under various values, and:
DivAccSet=[DivAcc1,DivAcc2,......,DivAcckn,......,DivAccKwNum]。
Every refund information both corresponds to a refund task, each task to be refunded include object to be refunded and to also
The money amount of money, wherein object to be refunded includes but is not limited to housing loan, vehicle loan, credit card, P2P loan etc..
In the present solution, only needing to filter out current refund week from the information aggregate according to the keyword set
Task to be refunded in phase.
Generally, the payment period can be one month, in initial borrow money, can by user's self-setting or
Last repayment date of the date as every month is arranged by Server Default in person, between two continuous finally repayment dates when
Between section be a complete payment period.For example, if the user setting 15 days are used as last repayment date, 16 days to 2 January
The moon 15 was a payment period, and 16 days 2 months to March 15 are a payment period ... ..., and so on.
Generally, before repayment date, the modes such as each bank can all be pushed by short message, message remind user to refund.It can
With by keyword from selected out in numerous short messages, PUSH message the user in current payment period wait refund appoint
Business.
Step S103, the task to be refunded is handled using preset neural network model, obtain to it is described to
The refund scheme of refund task.
As shown in Fig. 2, step S103 can specifically include following process:
Step S1031, obtaining to be refunded task and the user of the user in current payment period can be used for also
The account of each savings account of money.
The account for each savings account that the user can be used for refunding can voluntarily be added or deleted by the user, excellent
Selection of land needs user's authorization to the search access right of its account when adding new account.
In the present embodiment, the account for each savings account that user can be used for refunding can be stored in specified data library
Number, when closing on last repayment date, not yet receiving the refund of the user, then obtain what it can be used for refunding from the database
The account of each savings account.
Step S1032, determination bank server corresponding with the account of each savings account, and taken to each bank
Business device sends inquiry request.
In the present embodiment, can by inquiring the account of each savings account in the People's Bank's Internetbank inter-bank system, from
And determine corresponding each bank server, and can take by the People's Bank's Internetbank inter-bank system to each bank
Business device sends inquiry request, in the inquiry request, including going through in the remaining sum and designated time period to each savings account
The inquiry of history expenditure bill.
Step S1033, the response message of each bank server feedback is received, and is obtained respectively from the response message
The remaining sum of each savings account and the history in designated time period pay bill.
Bank server is after receiving the inquiry request, to refund management server namely execution master of the invention
Body, feedback response message include that the remaining sum of savings account and the history in designated time period pay bill in the response message.
Preferably, in order to guarantee safety, data are avoided to be utilized by criminal, bank server is receiving described look into
After asking request, authorization requests can also be sent to the terminal of the user, include refund management server in the authorization requests
Device identification, which can carry in the inquiry request, and the terminal of the user, which manages the refund, to be taken
The device identification of business device is checked, if confirmation, sends authorized order, the bank service to the bank server
Device is after receiving the authorized order, to the refund management server feedback response message.
Refund management server, can be from wherein obtaining more than each savings account after receiving the response message respectively
History in volume and designated time period pays bill.
Step S1034, history expenditure bill is handled using the neural network model, obtains each savings
Expectan volume of the account in next payment period.
Firstly, the disbursement of each payment period of c-th of the savings account of construction as follows in the designated time period
Sequence:
H={ h1,h2,...,ht,...,hT}
Wherein, c is the serial number of savings account, and 1≤c≤CardNum, CardNum are the storage that the user can be used for refunding
The sum of account is stored, 1≤t≤T, T are the length of the disbursement sequence, htFor the disbursement of t-th of payment period, and1≤n≤Nt, NtFor the expenditure total degree of t-th of payment period, ht_nFor the n-th branch of t-th of payment period
Number out.
Generally, the designated time period can be 1 year, then the length T=12 of the disbursement sequence.
Then, the neural network model is trained using the disbursement sequence, obtains trained nerve net
Network model.
Specifically, Artificial Neural Network Structures as shown in Figure 3 are used in the present embodiment, various are successively selected according to following
It takes the value in the disbursement sequence to be trained the neural network model, determines in the neural network model
Each neural unit parameter:
it=σ (ht-1Ui+xtWi)
ft=σ (ht-1Uf+xtWf)
ot=σ (ht-1Uo+xtWo)
ht=ottanh(ct)
Wherein, xtFor the sequence number of t-th of payment period, itFor the output of preset first unit, ftIt is single for preset second
Member output, otIt is exported for preset third unit,For the output of preset Unit the 4th, ctFor the output of preset Unit the 5th, σ
For sigmoid function, namelyUi、Uf、Uc、Wi、Wf、WcThe neural unit parameter respectively to be determined.
In the neural network model, entire treatment process is consisted of three parts: i.e. input gate processing, forget door processing and
Out gate processing.
Wherein, forget the output h that door is the above payment periodt-1With the input x of this payment periodtFor input
Sigmoid function is ct-1In each single item generate a value in [0,1], to control the memory state of a upper payment period
The degree to pass into silence.Input gate and a tanh functionWhich new information cooperation control has added
Enter.Tanh function generates a new candidate vectorInput gate isIn each single item generate a value in [0,1],
Control new information be added into number.There is the output f for forgeing doort, for controlling the degree that a upper payment period passes into silence,
There has also been the output i of input gatet, for the number that new information is added into controlled, so that it may update the memory of this payment period
State.Out gate be used to control this payment period memory state how many be filtered.First by this payment period
Memory state activation, out gate are that wherein each single item generates a value in [0,1], control the memory shape of this payment period
The degree that state is filtered.
As can be seen that the disbursement information of previous payment period is cached by the status architecture in structure shown in Fig. 3, and
And history disbursement information is safeguarded by input gate, forgetting door and out gate, to realize long range history letter
The valid cache of breath.
The neural network model is trained by using the disbursement sequence of construction, available each refund
The parameters such as the transitive relation of disbursement between the period, namely folding situation and U, W for obtaining each door.
Finally, calculating c-th of savings account in next payment period by the trained neural network model
Expectan volume.
Step S1035, according to the remaining sum of each savings account and the expectan volume in next payment period point
The available amount to pay of each savings account is not calculated.
For example, the available amount to pay of a certain savings account can be calculate by the following formula:
Account=Balance-Expenses
Wherein, Account is the available amount to pay of the savings account, and Balance is the remaining sum of the savings account,
Expenses is expectan volume of the savings account in next payment period.
Preferably, in order to guarantee certain surplus, the available amount to pay of the savings account can also be calculate by the following formula:
Account=Balance-Expenses-Margin
Wherein, Margin is preset margin value.
Step S1036, the refund scheme to the task to be refunded is determined according to the available amount to pay of each savings account.
The record firstly, history for obtaining the user is refunded, then refunds to record according to the history of the user and count respectively
Calculate the priority index that each savings account refunds to each task to be refunded.
For example, the preferential finger that each savings account refunds to each task to be refunded can be calculated separately according to the following formula
Number:
Wherein, c is the serial number of savings account, and 1≤c≤CardNum, CardNum are the storage that the user can be used for refunding
Store the sum of account, PayNuml,cFor the history refund user described in record using c-th savings account to first to
The number that refund task is refunded, PayPril,cC-th of savings account is used to refund first of task to be refunded
Priority index.
The preceding M savings account maximum to the priority index of first of task to be refunded for meeting following condition is determined as
The savings account refunded to first of task to be refunded:
AndWherein, l is the sequence of task to be refunded
Number, 1≤l≤LoanNum, LoanNum are the sum of task to be refunded, ActThreshlIt is first task to be refunded to also
The money amount of money, Accountl,mFor being arranged sequentially on the position m from big to small according to the priority index to first of task to be refunded
Savings account available amount to pay;
The amount to pay for each savings account refunded to first of task to be refunded is calculated according to the following formula:
Account′l,m=Accountl,m, 1≤m < M
Wherein, Account 'l,mFor being arranged sequentially from big to small according to the priority index to first of task to be refunded
The amount to pay of savings account on the position m.
Further, after determining to the refund scheme of the task to be refunded, the refund scheme can also include:
The refund scheme is sent in the terminal of the user, and receives the user by the terminal to described
The feedback information of refund scheme.
The user is after receiving the refund scheme, can be at the terminal to the refund scheme if meeting its expection
Confirmed, the feedback information of terminal is that confirmation message can be at the terminal to the refund scheme if not meeting its expection at this time
It modifies, for example, the savings account of modification refund or the amount to pay for modifying some savings account etc., obtain modified
Refund scheme, the feedback information of terminal is scheme modifying information at this time.
If the feedback information is confirmation message, refund according to the refund scheme;
If the feedback information is scheme modifying information, modified refund side in the scheme modifying information is extracted
Case, and refund according to the modified refund scheme.
In conclusion the embodiment of the present invention obtains the information aggregate sent by specified data source, the information collection first
Then information in conjunction including one or more is filtered out from the information aggregate wait refund according to preset keyword set
Task includes more than one keyword in the keyword set, finally, using preset neural network model to it is described to
Refund task is handled, and the refund scheme to the task to be refunded is obtained.Through the embodiment of the present invention, it may be implemented automatic
Task to be refunded is filtered out from massive information, and refund scheme is generated for user automatically, can avoid according to the refund scheme
Because carelessness causes to forget the case where refunding, user experience is greatly improved.
It should be understood that the size of the serial number of each step is not meant that the order of the execution order in above-described embodiment, each process
Execution sequence should be determined by its function and internal logic, the implementation process without coping with the embodiment of the present invention constitutes any limit
It is fixed.
Corresponding to a kind of repayment method described in foregoing embodiments, Fig. 4 shows one kind provided in an embodiment of the present invention also
One embodiment structure chart of money device.
In the present embodiment, a kind of refund device may include:
Information aggregate obtains module 401, for obtaining the information aggregate sent by specified data source, the information aggregate
In include one or more information;
Task screening module 402 to be refunded, for being filtered out from the information aggregate according to preset keyword set
Task to be refunded includes more than one keyword in the keyword set;
Model processing modules 403 are obtained for being handled using preset neural network model the task to be refunded
To the refund scheme to the task to be refunded.
Further, the refund device can also include:
Building of corpus module, for constructing the first corpus and the second language according to the historical information in preset database
Expect library, include each history refund information in first corpus, includes removing history refund information in second corpus
Except each other historical informations;
Word frequency statistics module, the frequency occurred in first corpus for counting each candidate word respectively
The frequency that secondary and each candidate word occurs in second corpus, candidate's word is in first corpus
The word occurred in library;
Discrimination computing module, for calculating separately the discrimination of each candidate word according to the following formula:
Wherein, w is the serial number of each candidate word, and 1≤w≤WordNum, WordNum are the sum of candidate word,
FstNumwFor the frequency that w-th of candidate word occurs in first corpus, SndNumwIt is w-th of candidate word in institute
State the frequency occurred in the second corpus, DivDegwFor the discrimination of w-th of candidate word.
Keyword set constructing module, for selecting keyword from each candidate word according to the discrimination, and
The keyword selected is configured to the keyword set.
Further, the model processing modules may include:
Information acquisition unit, can for obtaining to be refunded task and the user of the user in current payment period
The account of each savings account for refund;
Inquiry request transmission unit, for determining bank server corresponding with the account of each savings account, and
Inquiry request is sent to each bank server;
Response message receiving unit disappears for receiving the response message of each bank server feedback, and from the response
The remaining sum of each savings account and the history expenditure bill in designated time period are obtained in breath respectively;
Expectan volume computing unit, for being paid at bill using the neural network model to the history
Reason, obtains expectan volume of each savings account in next payment period;
Amount to pay computing unit can be used, for the remaining sum according to each savings account and in next payment period
Expectan volume calculates separately the available amount to pay of each savings account;
Refund scheme determination unit, for being determined according to the available amount to pay of each savings account to the task to be refunded
Refund scheme.
Further, the expectan volume computing unit may include:
Disbursement sequence structure subelement, for constructing c-th following of savings account in the designated time period
The disbursement sequence of each payment period:
H={ h1,h2,...,ht,...,hT}
Wherein, c is the serial number of savings account, and 1≤c≤CardNum, CardNum are the storage that the user can be used for refunding
The sum of account is stored, 1≤t≤T, T are the length of the disbursement sequence, htFor the disbursement of t-th of payment period, and1≤n≤Nt, NtFor the expenditure total degree of t-th of payment period, ht_nFor the n-th branch of t-th of payment period
Number out;
Model training subelement is obtained for being trained using the disbursement sequence to the neural network model
Trained neural network model;
Expectan volume computation subunit, for calculating c-th of savings account by the trained neural network model
Expectan volume of the family in next payment period.
Further, the model training subelement is specifically used for successively choosing the disbursement sequence according to as follows various
Value in column is trained the neural network model, determines each neural unit ginseng in the neural network model
Number:
it=σ (ht-1Ui+xtWi)
ft=σ (ht-1Uf+xtWf)
ot=σ (ht-1Uo+xtWo)
ht=ottanh(ct)
Wherein, xtFor the sequence number of t-th of payment period, itFor the output of preset first unit, ftIt is single for preset second
Member output, otIt is exported for preset third unit,For the output of preset Unit the 4th, ctFor the output of preset Unit the 5th, σ
For sigmoid function, Ui、Uf、Uc、Wi、Wf、WcThe neural unit parameter respectively to be determined.
Further, the refund scheme determination unit may include:
Record of refunding obtains subelement, and the history for obtaining the user, which is refunded, to be recorded;
It is preferred that index computation subunit, calculates separately each savings account for refunding to record according to the history of the user
The priority index refunded to each task to be refunded;
Refund account determines subelement, for that will meet the priority index to first of task to be refunded of following condition most
Big preceding M savings account is determined as the savings account refunded to first of task to be refunded:
AndWherein, l is the sequence of task to be refunded
Number, 1≤l≤LoanNum, LoanNum are the sum of task to be refunded, ActThreshlIt is first task to be refunded to also
The money amount of money, Accountl,mFor being arranged sequentially on the position m from big to small according to the priority index to first of task to be refunded
Savings account available amount to pay;
Amount to pay determines subelement, for calculating each savings refunded to first of task to be refunded according to the following formula
The amount to pay of account:
Account′l,m=Accountl,m, 1≤m < M
Wherein, Account 'l,mFor being arranged sequentially from big to small according to the priority index to first of task to be refunded
The amount to pay of savings account on the position m.
Further, the preferred index computation subunit is specifically used for calculating separately each savings account pair according to the following formula
The priority index that each task to be refunded is refunded:
Wherein, c is the serial number of savings account, and 1≤c≤CardNum, CardNum are the storage that the user can be used for refunding
Store the sum of account, PayNuml,cFor the history refund user described in record using c-th savings account to first to
The number that refund task is refunded, PayPril,cC-th of savings account is used to refund first of task to be refunded
Priority index.
It is apparent to those skilled in the art that for convenience and simplicity of description, the device of foregoing description,
The specific work process of module and unit, can refer to corresponding processes in the foregoing method embodiment, and details are not described herein.
In the above-described embodiments, it all emphasizes particularly on different fields to the description of each embodiment, is not described in detail or remembers in some embodiment
The part of load may refer to the associated description of other embodiments.
The schematic block diagram that Fig. 5 shows a kind of server provided in an embodiment of the present invention illustrates only for ease of description
Part related to the embodiment of the present invention.
In the present embodiment, the server 5 may include: processor 50, memory 51 and be stored in the storage
In device 51 and the computer-readable instruction 52 that can run on the processor 50, such as execute the calculating of above-mentioned repayment method
Machine readable instruction.The processor 50 is realized when executing the computer-readable instruction 52 in above-mentioned each repayment method embodiment
The step of, such as step S101 to S103 shown in FIG. 1.Alternatively, the processor 50 executes the computer-readable instruction 52
The function of each module/unit in the above-mentioned each Installation practice of Shi Shixian, such as the function of module 401 to 403 shown in Fig. 4.
Illustratively, the computer-readable instruction 52 can be divided into one or more module/units, one
Or multiple module/units are stored in the memory 51, and are executed by the processor 50, to complete the present invention.Institute
Stating one or more module/units can be the series of computation machine readable instruction section that can complete specific function, the instruction segment
For describing implementation procedure of the computer-readable instruction 52 in the server 5.
The processor 50 can be central processing unit (Central Processing Unit, CPU), can also be
Other general processors, digital signal processor (Digital Signal Processor, DSP), specific integrated circuit
(Application Specific Integrated Circuit, ASIC), field programmable gate array (Field-
Programmable Gate Array, FPGA) either other programmable logic device, discrete gate or transistor logic,
Discrete hardware components etc..General processor can be microprocessor or the processor is also possible to any conventional processor
Deng.
The memory 51 can be the internal storage unit of the server 5, such as the hard disk or memory of server 5.
The memory 51 is also possible to the External memory equipment of the server 5, such as the plug-in type being equipped on the server 5 is hard
Disk, intelligent memory card (Smart Media Card, SMC), secure digital (Secure Digital, SD) card, flash card
(Flash Card) etc..Further, the memory 51 can also both include the internal storage unit of the server 5 or wrap
Include External memory equipment.The memory 51 is for storing needed for the computer-readable instruction and the server 5 it
Its instruction and data.The memory 51 can be also used for temporarily storing the data that has exported or will export.
The functional units in various embodiments of the present invention may be integrated into one processing unit, is also possible to each
Unit physically exists alone, and can also be integrated in one unit with two or more units.Above-mentioned integrated unit both may be used
To use formal implementation of hardware, can also realize in the form of software functional units.
If the integrated unit is realized in the form of SFU software functional unit and sells or use as independent product
When, it can store in a computer readable storage medium.Based on this understanding, technical solution of the present invention substantially or
Person says that all or part of the part that contributes to existing technology or the technical solution can body in the form of software products
Reveal and, which is stored in a storage medium, including several computer-readable instructions are used so that one
Platform computer equipment (can be personal computer, server or the network equipment etc.) executes described in each embodiment of the present invention
The all or part of the steps of method.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory (ROM, Read-
Only Memory), random access memory (RAM, Random Access Memory), magnetic or disk etc. are various can be with
Store the medium of computer-readable instruction.
Embodiment described above is merely illustrative of the technical solution of the present invention, rather than its limitations;Although referring to aforementioned reality
Applying example, invention is explained in detail, those skilled in the art should understand that: it still can be to aforementioned each
Technical solution documented by embodiment is modified or equivalent replacement of some of the technical features;And these are modified
Or replacement, the spirit and scope for technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution.
Claims (10)
1. a kind of repayment method neural network based characterized by comprising
The information aggregate sent by specified data source is obtained, includes the information of one or more in the information aggregate;
Task to be refunded is filtered out from the information aggregate according to preset keyword set, includes in the keyword set
More than one keyword;
The task to be refunded is handled using preset neural network model, obtains the refund to the task to be refunded
Scheme.
2. repayment method according to claim 1, which is characterized in that the setting up procedure of the keyword set includes:
The first corpus and the second corpus are constructed according to the historical information in preset database, is wrapped in first corpus
Each history refund information is included, includes each other historical informations in addition to history refund information in second corpus;
The frequency that each candidate word occurs in first corpus and each candidate word are counted respectively described the
The frequency occurred in two corpus, candidate's word is the word occurred in first corpus;
The discrimination of each candidate word is calculated separately according to the following formula:
Wherein, w is the serial number of each candidate word, and 1≤w≤WordNum, WordNum are the sum of candidate word, FstNumwFor
The frequency that w-th of candidate word occurs in first corpus, SndNumwIt is w-th of candidate word in second language
The frequency occurred in material library, DivDegwFor the discrimination of w-th of candidate word.
Keyword is selected from each candidate word according to the discrimination, and the keyword selected is configured to the pass
Keyword set.
3. repayment method according to claim 1, which is characterized in that described to use preset neural network model to described
Task to be refunded is handled, obtain include: to the refund scheme of the task to be refunded
Obtain each savings account that be refunded task and the user of the user in current payment period can be used for refunding
The account at family;
It determines bank server corresponding with the account of each savings account, and sends inquiry to each bank server and ask
It asks;
The response message of each bank server feedback is received, and obtains each savings account respectively from the response message
History in remaining sum and designated time period pays bill;
History expenditure bill is handled using the neural network model, each savings account is obtained and is gone back next
Expectan volume in the money period;
Each savings are calculated separately according to the remaining sum of each savings account and the expectan volume in next payment period
The available amount to pay of account;
The refund scheme to the task to be refunded is determined according to the available amount to pay of each savings account.
4. repayment method according to claim 3, which is characterized in that the available amount to pay according to each savings account
Determination includes: to the refund scheme of the task to be refunded
The history for obtaining the user, which is refunded, to be recorded;
It is refunded to record according to the history of the user and calculates separately what each savings account refunded to each task to be refunded
Priority index;
The preceding M savings account maximum to the priority index of first of task to be refunded for meeting following condition is determined as to l
The savings account that a task to be refunded is refunded:
And
Wherein, l is the serial number of task to be refunded, and 1≤l≤LoanNum, LoanNum are the sum of task to be refunded,
ActThreshlIt is first task to be refunded to repayment amount, Accountl,mFor according to the excellent of first task to be refunded
The available amount to pay of the savings account being arranged sequentially on the position m of first index from big to small;
The amount to pay for each savings account refunded to first of task to be refunded is calculated according to the following formula:
Account′l,m=Accountl,m, 1≤m < M
Wherein, Account 'l,mFor according to the priority index to first of task to be refunded from big to small be arranged sequentially m
On savings account amount to pay.
5. repayment method according to claim 4, which is characterized in that described according to the history of user refund record point
Not calculating the priority index that each savings account refunds to each task to be refunded includes:
The priority index that each savings account refunds to each task to be refunded is calculated separately according to the following formula:
Wherein, c is the serial number of savings account, and 1≤c≤CardNum, CardNum are the savings account that the user can be used for refunding
The sum at family, PayNuml,cUse c-th of savings account to first wait refund for the user described in history refund record
The number that task is refunded, PayPril,cIt is preferential to use c-th of savings account to refund first of task to be refunded
Index.
6. a kind of computer readable storage medium, the computer-readable recording medium storage has computer-readable instruction, special
Sign is, the refund side as described in any one of claims 1 to 5 is realized when the computer-readable instruction is executed by processor
The step of method.
7. a kind of server, including memory, processor and storage can transport in the memory and on the processor
Capable computer-readable instruction, which is characterized in that the processor realizes following steps when executing the computer-readable instruction:
The information aggregate sent by specified data source is obtained, includes the information of one or more in the information aggregate;
Task to be refunded is filtered out from the information aggregate according to preset keyword set, includes in the keyword set
More than one keyword;
The task to be refunded is handled using preset neural network model, obtains the refund to the task to be refunded
Scheme.
8. server according to claim 7, which is characterized in that the setting up procedure of the keyword set includes:
The first corpus and the second corpus are constructed according to the historical information in preset database, is wrapped in first corpus
Each history refund information is included, includes each other historical informations in addition to history refund information in second corpus;
The frequency that each candidate word occurs in first corpus and each candidate word are counted respectively described the
The frequency occurred in two corpus, candidate's word is the word occurred in first corpus;
The discrimination of each candidate word is calculated separately according to the following formula:
Wherein, w is the serial number of each candidate word, and 1≤w≤WordNum, WordNum are the sum of candidate word, FstNumwFor
The frequency that w-th of candidate word occurs in first corpus, SndNumwIt is w-th of candidate word in second language
The frequency occurred in material library, DivDegwFor the discrimination of w-th of candidate word.
Keyword is selected from each candidate word according to the discrimination, and the keyword selected is configured to the pass
Keyword set.
9. server according to claim 7, which is characterized in that it is described using preset neural network model to it is described to
Refund task is handled, obtain include: to the refund scheme of the task to be refunded
Obtain each savings account that be refunded task and the user of the user in current payment period can be used for refunding
The account at family;
It determines bank server corresponding with the account of each savings account, and sends inquiry to each bank server and ask
It asks;
The response message of each bank server feedback is received, and obtains each savings account respectively from the response message
History in remaining sum and designated time period pays bill;
History expenditure bill is handled using the neural network model, each savings account is obtained and is gone back next
Expectan volume in the money period;
Each savings are calculated separately according to the remaining sum of each savings account and the expectan volume in next payment period
The available amount to pay of account;
The refund scheme to the task to be refunded is determined according to the available amount to pay of each savings account.
10. server according to claim 9, which is characterized in that the available amount to pay according to each savings account
Determination includes: to the refund scheme of the task to be refunded
The history for obtaining the user, which is refunded, to be recorded;
It is refunded to record according to the history of the user and calculates separately what each savings account refunded to each task to be refunded
Priority index;
The preceding M savings account maximum to the priority index of first of task to be refunded for meeting following condition is determined as to l
The savings account that a task to be refunded is refunded:
And
Wherein, l is the serial number of task to be refunded, and 1≤l≤LoanNum, LoanNum are the sum of task to be refunded,
ActThreshlIt is first task to be refunded to repayment amount, Accountl,mFor according to the excellent of first task to be refunded
The available amount to pay of the savings account being arranged sequentially on the position m of first index from big to small;
The amount to pay for each savings account refunded to first of task to be refunded is calculated according to the following formula:
Account′l,m=Accountl,m, 1≤m < M
Wherein, Account 'l,mFor according to the priority index to first of task to be refunded from big to small be arranged sequentially m
On savings account amount to pay.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811533834.3A CN109785112A (en) | 2018-12-14 | 2018-12-14 | Repayment method, computer readable storage medium and server neural network based |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811533834.3A CN109785112A (en) | 2018-12-14 | 2018-12-14 | Repayment method, computer readable storage medium and server neural network based |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109785112A true CN109785112A (en) | 2019-05-21 |
Family
ID=66496976
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811533834.3A Pending CN109785112A (en) | 2018-12-14 | 2018-12-14 | Repayment method, computer readable storage medium and server neural network based |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109785112A (en) |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20020052863A1 (en) * | 2000-10-31 | 2002-05-02 | Tonfu Corporation | Law retrieval system, law retrieval apparatus and law retrieval program |
CN105447750A (en) * | 2015-11-17 | 2016-03-30 | 小米科技有限责任公司 | Information identification method, apparatus, terminal and server |
CN105989040A (en) * | 2015-02-03 | 2016-10-05 | 阿里巴巴集团控股有限公司 | Intelligent question-answer method, device and system |
CN108492170A (en) * | 2018-02-26 | 2018-09-04 | 平安普惠企业管理有限公司 | A kind of repayment method, computer readable storage medium and terminal device |
CN108574732A (en) * | 2018-03-22 | 2018-09-25 | 平安科技(深圳)有限公司 | A kind of PUSH message processing method, computer readable storage medium and terminal device |
-
2018
- 2018-12-14 CN CN201811533834.3A patent/CN109785112A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20020052863A1 (en) * | 2000-10-31 | 2002-05-02 | Tonfu Corporation | Law retrieval system, law retrieval apparatus and law retrieval program |
CN105989040A (en) * | 2015-02-03 | 2016-10-05 | 阿里巴巴集团控股有限公司 | Intelligent question-answer method, device and system |
CN105447750A (en) * | 2015-11-17 | 2016-03-30 | 小米科技有限责任公司 | Information identification method, apparatus, terminal and server |
CN108492170A (en) * | 2018-02-26 | 2018-09-04 | 平安普惠企业管理有限公司 | A kind of repayment method, computer readable storage medium and terminal device |
CN108574732A (en) * | 2018-03-22 | 2018-09-25 | 平安科技(深圳)有限公司 | A kind of PUSH message processing method, computer readable storage medium and terminal device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10417556B1 (en) | Simulation-based controls optimization using time series data forecast | |
CN108492170A (en) | A kind of repayment method, computer readable storage medium and terminal device | |
US5991745A (en) | Reverse mortgage loan calculation system and process | |
US20200380481A1 (en) | Autonomous bill pay bot with automatic social overdraft protection | |
CN109697667B (en) | Quota management and control method and device based on trust mode | |
JP6815974B2 (en) | Balance management system | |
CN113657901B (en) | Method, system, terminal and medium for managing fee owed users | |
US11514535B2 (en) | Method and apparatus for the optimization of assets and an asset allocation register for personal use, enterprise or business use, and savings and retirement | |
CN107392614B (en) | Method and device for realizing offline transaction | |
CN109919465A (en) | Financial risk early-warning method and apparatus | |
CN109740792A (en) | Data predication method, system, terminal and computer storage medium | |
CN110348981A (en) | Loan air control method and device, electronic equipment and computer readable storage medium | |
CN112927064A (en) | Deferred payment data processing method and device, electronic equipment and storage medium | |
CN110930252A (en) | Digital currency transaction data processing method, device and system | |
CN109785112A (en) | Repayment method, computer readable storage medium and server neural network based | |
CN110245985A (en) | A kind of information processing method and device | |
CN107305662A (en) | Recognize the method and device of violation account | |
CN108596751A (en) | A kind of account repayment method, device, terminal device and storage medium | |
CN116071167A (en) | Object anomaly prediction method, device and computer readable storage medium | |
US20240087030A1 (en) | Method And System For Performing Merchant Cash Advances | |
Arora et al. | A Hybrid Neuro-Fuzzy Network for Underwriting of Life Insurance. | |
KR102415826B1 (en) | Method for providing automatic collection service for moneybox | |
CN111967795B (en) | Cash self-circulation method and device for banking outlets, storage medium and electronic equipment | |
CN112116169B (en) | User behavior determining method and device and electronic equipment | |
US20240144388A1 (en) | System and method for calculating and disbursing advanced wages |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190521 |
|
WD01 | Invention patent application deemed withdrawn after publication |